chore(release): Update version to v1.4.349

Merge pull request #1877 from ksylvan/kayvan/modernize-part4-string-and-slice-syntax
modernize: update GitHub Actions and modernize Go code
2026-01-09 22:38:10 -05:00 · 2025-12-16 08:12:11 +00:00 · 2025-12-16 00:09:43 -08:00 · 2025-12-16 00:06:39 -08:00 · 2025-12-15 23:55:37 -08:00 · 2025-12-16 07:34:45 +00:00
110 changed files with 4734 additions and 1849 deletions
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -20,18 +20,22 @@ jobs:
      contents: read
    steps:
      - name: Checkout code
-        uses: actions/checkout@v5
-
-      - name: Install Nix
-        uses: DeterminateSystems/nix-installer-action@main
+        uses: actions/checkout@v6

      - name: Set up Go
-        uses: actions/setup-go@v5
+        uses: actions/setup-go@v6
        with:
          go-version-file: ./go.mod

      - name: Run tests
        run: go test -v ./...

+      - name: Check for modernization opportunities
+        run: |
+          go run golang.org/x/tools/go/analysis/passes/modernize/cmd/modernize@latest ./...
+
+      - name: Install Nix
+        uses: DeterminateSystems/nix-installer-action@v21
+
      - name: Check Formatting
        run: nix flake check
--- a/.github/workflows/patterns.yaml
+++ b/.github/workflows/patterns.yaml
@@ -11,7 +11,7 @@ jobs:
    runs-on: ubuntu-latest
    steps:
      - name: Checkout code
-        uses: actions/checkout@v5
+        uses: actions/checkout@v6
        with:
          fetch-depth: 0

@@ -32,7 +32,7 @@ jobs:

      - name: Upload Patterns Artifact
        if: steps.check-changes.outputs.changes == 'true'
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@v6
        with:
          name: patterns
          path: patterns.zip
--- a/.github/workflows/release.yml
+++ b/.github/workflows/release.yml
@@ -15,12 +15,12 @@ jobs:
      contents: read
    steps:
      - name: Checkout code
-        uses: actions/checkout@v5
+        uses: actions/checkout@v6
        with:
          fetch-depth: 0

      - name: Set up Go
-        uses: actions/setup-go@v5
+        uses: actions/setup-go@v6
        with:
          go-version-file: ./go.mod

@@ -37,11 +37,11 @@ jobs:
      contents: write
    steps:
      - name: Checkout code
-        uses: actions/checkout@v5
+        uses: actions/checkout@v6
        with:
          fetch-depth: 0
      - name: Set up Go
-        uses: actions/setup-go@v5
+        uses: actions/setup-go@v6
        with:
          go-version-file: ./go.mod
      - name: Run GoReleaser
--- a/.github/workflows/update-version-and-create-tag.yml
+++ b/.github/workflows/update-version-and-create-tag.yml
@@ -24,17 +24,17 @@ concurrency:
 jobs:
  update-version:
    if: >
-      ${{ github.repository_owner == 'danielmiessler' }} &&
+      github.repository_owner == 'danielmiessler' &&
      github.event_name == 'push' && github.ref == 'refs/heads/main'
    runs-on: ubuntu-latest
    steps:
      - name: Checkout repository
-        uses: actions/checkout@v5
+        uses: actions/checkout@v6
        with:
          fetch-depth: 0

      - name: Install Nix
-        uses: DeterminateSystems/nix-installer-action@main
+        uses: DeterminateSystems/nix-installer-action@v21

      - name: Set up Git
        run: |
--- a/.vscode/settings.json
+++ b/.vscode/settings.json
@@ -10,11 +10,13 @@
 		"aplicar",
 		"atotto",
 		"Autonoe",
+		"azureml",
 		"badfile",
 		"Behrens",
 		"blindspots",
 		"Bombal",
 		"Buildx",
+		"byid",
 		"Callirhoe",
 		"Callirrhoe",
 		"Cerebras",
@@ -22,6 +24,7 @@
 		"compadd",
 		"compdef",
 		"compinit",
+		"conceptmap",
 		"creatordate",
 		"curcontext",
 		"custompatterns",
@@ -86,12 +89,14 @@
 		"horts",
 		"HTMLURL",
 		"imagetools",
+		"Jamba",
 		"jaredmontoya",
 		"jessevdk",
 		"Jina",
 		"joho",
 		"kballard",
 		"Keploy",
+		"kimi",
 		"Kore",
 		"ksylvan",
 		"Langdock",
@@ -147,6 +152,7 @@
 		"Pulcherrima",
 		"pycache",
 		"pyperclip",
+		"qwen",
 		"readystream",
 		"restapi",
 		"rmextension",
@@ -160,6 +166,7 @@
 		"sess",
 		"sgaunet",
 		"shellquote",
+		"skeletonlabs",
 		"SSEHTTP",
 		"storer",
 		"Streamlit",
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,332 @@
 # Changelog

+## v1.4.349 (2025-12-16)
+
+### PR [#1877](https://github.com/danielmiessler/Fabric/pull/1877) by [ksylvan](https://github.com/ksylvan): modernize: update GitHub Actions and modernize Go code
+
+- Modernize GitHub Actions and Go code with latest stdlib features
+- Upgrade GitHub Actions to latest versions (v6, v21) and add modernization check step
+- Replace strings manipulation with `strings.CutPrefix` and `strings.CutSuffix`
+- Replace manual loops with `slices.Contains` for validation and use `strings.SplitSeq` for iterator-based splitting
+- Replace `fmt.Sprintf` with `fmt.Appendf` for efficiency and simplify padding calculation with `max` builtin
+
+## v1.4.348 (2025-12-16)
+
+### PR [#1876](https://github.com/danielmiessler/Fabric/pull/1876) by [ksylvan](https://github.com/ksylvan): modernize Go code with TypeFor and range loops
+
+- Replace reflect.TypeOf with TypeFor generic syntax for improved type handling
+- Convert traditional for loops to range-based iterations for better code readability
+- Simplify reflection usage in CLI flag handling to reduce complexity
+- Update test loops to use range over integers for cleaner test code
+- Refactor string processing loops in template plugin to use modern Go patterns
+
+## v1.4.347 (2025-12-16)
+
+### PR [#1875](https://github.com/danielmiessler/Fabric/pull/1875) by [ksylvan](https://github.com/ksylvan): modernize: update benchmarks to use b.Loop and refactor map copying
+
+- Updated benchmark loops to use cleaner `b.Loop()` syntax
+- Removed unnecessary `b.ResetTimer()` call in token benchmark
+- Used `maps.Copy` for merging variables in patterns handler
+
+## v1.4.346 (2025-12-16)
+
+### PR [#1874](https://github.com/danielmiessler/Fabric/pull/1874) by [ksylvan](https://github.com/ksylvan): refactor: replace interface{} with any across codebase
+
+- Part 1 of dealing with #1873 as pointed out by @philoserf
+- Replace `interface{}` with `any` in slice type declarations throughout the codebase
+- Update map types from `map[string]interface{}` to `map[string]any` for modern Go standards
+- Change variadic function parameters to use `...any` instead of `...interface{}`
+- Modernize JSON unmarshaling variables to use `any` for consistency
+- Update struct fields and method signatures to prefer the `any` alias over legacy interface syntax
+
+## v1.4.345 (2025-12-15)
+
+### PR [#1870](https://github.com/danielmiessler/Fabric/pull/1870) by [ksylvan](https://github.com/ksylvan): Web UI: upgrade pdfjs and add SSR-safe dynamic PDF worker init
+
+- Upgrade `pdfjs-dist` to v5 with new engine requirement
+- Dynamically import PDF.js to avoid SSR import-time crashes
+- Configure PDF worker via CDN using runtime PDF.js version
+- Update PDF conversion pipeline to use lazy initialization
+- Guard chat message localStorage persistence behind browser checks
+
+## v1.4.344 (2025-12-14)
+
+### PR [#1867](https://github.com/danielmiessler/Fabric/pull/1867) by [jaredmontoya](https://github.com/jaredmontoya): chore: update flake
+
+- Chore: update flake
+- Merge branch 'main' into update-flake
+
+## v1.4.343 (2025-12-14)
+
+### PR [#1829](https://github.com/danielmiessler/Fabric/pull/1829) by [dependabo](https://github.com/apps/dependabot): chore(deps): bump js-yaml from 4.1.0 to 4.1.1 in /web in the npm_and_yarn group across 1 directory
+
+- Updated js-yaml dependency from version 4.1.0 to 4.1.1 in the /web directory
+
+## v1.4.342 (2025-12-13)
+
+### PR [#1866](https://github.com/danielmiessler/Fabric/pull/1866) by [ksylvan](https://github.com/ksylvan): fix: write CLI and streaming errors to stderr
+
+- Fix: write CLI and streaming errors to stderr
+- Route CLI execution errors to standard error output
+- Print Anthropic stream errors to stderr consistently
+- Add os import to support stderr error writes
+- Preserve help-output suppression and exit behavior
+
+## v1.4.341 (2025-12-10)
+
+### PR [#1860](https://github.com/danielmiessler/Fabric/pull/1860) by [ksylvan](https://github.com/ksylvan): fix: allow resetting required settings without validation errors
+
+- Fix: allow resetting required settings without validation errors
+- Update `Ask` to detect reset command and bypass validation
+- Refactor `OnAnswer` to support new `isReset` parameter logic
+- Invoke `ConfigureCustom` in `Setup` to avoid redundant re-validation
+- Add unit tests ensuring required fields can be reset
+
+## v1.4.340 (2025-12-08)
+
+### PR [#1856](https://github.com/danielmiessler/Fabric/pull/1856) by [ksylvan](https://github.com/ksylvan): Add support for new ClaudeHaiku 4.5 models
+
+- Add support for new ClaudeHaiku models in client
+- Add `ModelClaudeHaiku4_5` to supported models
+- Add `ModelClaudeHaiku4_5_20251001` to supported models
+
+## v1.4.339 (2025-12-08)
+
+### PR [#1855](https://github.com/danielmiessler/Fabric/pull/1855) by [ksylvan](https://github.com/ksylvan): feat: add image attachment support for Ollama vision models
+
+- Add multi-modal image support to Ollama client
+- Implement convertMessage to handle multi-content chat messages
+- Add loadImageBytes to fetch images from URLs
+- Support base64 data URLs for inline images
+- Handle HTTP image URLs with context propagation
+
+## v1.4.338 (2025-12-04)
+
+### PR [#1852](https://github.com/danielmiessler/Fabric/pull/1852) by [ksylvan](https://github.com/ksylvan): Add Abacus vendor for ChatLLM models with static model list
+
+- Add static model support and register Abacus provider
+- Detect modelsURL starting with 'static:' and route appropriately
+- Implement getStaticModels returning curated Abacus model list
+- Register Abacus provider with ModelsURL 'static:abacus'
+- Extend provider tests to include Abacus existence
+
+## v1.4.337 (2025-12-04)
+
+### PR [#1851](https://github.com/danielmiessler/Fabric/pull/1851) by [ksylvan](https://github.com/ksylvan): Add Z AI provider and glm model support
+
+- Add Z AI provider configuration to ProviderMap
+- Include BaseURL for Z AI API endpoint
+- Add test case for Z AI provider existence
+- Add glm to OpenAI model prefixes list
+- Support new Z AI provider in OpenAI compatible plugins
+
+## v1.4.336 (2025-12-01)
+
+### PR [#1848](https://github.com/danielmiessler/Fabric/pull/1848) by [zeddy303](https://github.com/zeddy303): Fix localStorage SSR error in favorites-store
+
+- Fix localStorage SSR error in favorites-store by using SvelteKit's browser constant instead of typeof localStorage check to properly handle server-side rendering and prevent 'localStorage.getItem is not a function' error when running dev server
+
+## v1.4.335 (2025-11-28)
+
+### PR [#1847](https://github.com/danielmiessler/Fabric/pull/1847) by [ksylvan](https://github.com/ksylvan): Improve model name matching for NeedsRaw in Ollama plugin
+
+- Improved model name matching in Ollama plugin by replacing prefix-based matching with substring matching
+- Enhanced NeedsRaw functionality to support more flexible model name detection
+- Renamed `ollamaPrefixes` variable to `ollamaSearchStrings` for better code clarity
+- Replaced `HasPrefix` function with `Contains` for more comprehensive model matching
+- Added "conceptmap" to VSCode dictionary settings
+
+### Direct commits
+
+- Merge branch 'danielmiessler:main' into main
+- Docs: Fix typo in README
+
+## v1.4.334 (2025-11-26)
+
+### PR [#1845](https://github.com/danielmiessler/Fabric/pull/1845) by [ksylvan](https://github.com/ksylvan): Add Claude Opus 4.5 Support
+
+- Add Claude Opus 4.5 model variants to Anthropic client
+- Upgrade anthropic-sdk-go from v1.16.0 to v1.19.0
+- Update golang.org/x/crypto from v0.41.0 to v0.45.0
+- Upgrade golang.org/x/net from v0.43.0 to v0.47.0
+- Bump golang.org/x/text from v0.28.0 to v0.31.0
+
+## v1.4.333 (2025-11-25)
+
+### PR [#1833](https://github.com/danielmiessler/Fabric/pull/1833) by [junaid18183](https://github.com/junaid18183): Added concall_summary
+
+- Added concall_summery pattern to extract strategic insights from earnings transcripts for investors.
+
+### PR [#1844](https://github.com/danielmiessler/Fabric/pull/1844) by [ksylvan](https://github.com/ksylvan): Correct directory name from `concall_summery` to `concall_summary`
+
+- Fix: correct directory name from `concall_summery` to `concall_summary`
+- Rename pattern directory to fix spelling error
+- Update suggest_pattern system with concall_summary references
+- Add concall_summary to BUSINESS and SUMMARIZE category listings
+- Add user documentation for earnings call analysis
+
+## v1.4.332 (2025-11-24)
+
+### PR [#1843](https://github.com/danielmiessler/Fabric/pull/1843) by [ksylvan](https://github.com/ksylvan): Implement case-insensitive vendor and model name matching
+
+- Fix: implement case-insensitive vendor and model name matching across the application
+- Add case-insensitive vendor lookup in VendorsManager
+- Implement model name normalization in GetChatter method
+- Add FilterByVendor method with case-insensitive matching
+- Add FindModelNameCaseInsensitive helper for model queries
+
+## v1.4.331 (2025-11-22)
+
+### PR [#1839](https://github.com/danielmiessler/Fabric/pull/1839) by [ksylvan](https://github.com/ksylvan): Add GitHub Models Provider and Refactor Fetching Fallback Logic
+
+- Add GitHub Models provider and refactor model fetching with direct API fallback
+- Add GitHub Models to supported OpenAI-compatible providers list
+- Implement direct HTTP fallback for non-standard model responses
+- Centralize model fetching logic in openai package
+- Upgrade openai-go SDK dependency from v1.8.2 to v1.12.0
+
+## v1.4.330 (2025-11-23)
+
+### PR [#1840](https://github.com/danielmiessler/Fabric/pull/1840) by [ZackaryWelch](https://github.com/ZackaryWelch): Replace deprecated bash function in completion script
+
+- Replace deprecated bash function in completion script to use `_comp_get_words` instead of `__get_comp_words_by_ref`, fixing compatibility issues with latest bash versions and preventing script breakage on updated distributions like Fedora 42+
+
+## v1.4.329 (2025-11-20)
+
+### PR [#1838](https://github.com/danielmiessler/fabric/pull/1838) by [ksylvan](https://github.com/ksylvan): refactor: implement i18n support for YouTube tool error messages
+
+- Replace hardcoded error strings with i18n translation calls
+- Add localization keys for YouTube errors to all locale files
+- Introduce `extractAndValidateVideoId` helper to reduce code duplication
+- Update timestamp parsing logic to handle localized error formats
+- Standardize error handling in `yt-dlp` execution with i18n
+
+## v1.4.328 (2025-11-18)
+
+### PR [#1836](https://github.com/danielmiessler/Fabric/pull/1836) by [ksylvan](https://github.com/ksylvan): docs: clarify `--raw` flag behavior for OpenAI and Anthropic providers
+
+- Update `--raw` flag description across all documentation files
+- Clarify flag only affects OpenAI-compatible providers behavior
+- Document Anthropic models use smart parameter selection
+- Remove outdated reference to system/user role changes
+- Update help text in CLI flags definition
+
+## v1.4.327 (2025-11-16)
+
+### PR [#1831](https://github.com/danielmiessler/Fabric/pull/1831) by [ksylvan](https://github.com/ksylvan): Remove `get_youtube_rss` pattern
+
+- Chore: remove `get_youtube_rss` pattern from multiple files
+- Remove `get_youtube_rss` from `pattern_explanations.md`
+- Delete `get_youtube_rss` entry in `pattern_descriptions.json`
+- Delete `get_youtube_rss` entry in `pattern_extracts.json`
+- Remove `get_youtube_rss` from `suggest_pattern/system.md`
+
+### PR [#1832](https://github.com/danielmiessler/Fabric/pull/1832) by [ksylvan](https://github.com/ksylvan): Improve channel management in Gemini provider
+
+- Fix: improve channel management in Gemini streaming method
+- Add deferred channel close at function start
+- Return error immediately instead of breaking loop
+- Remove redundant channel close statements from loop
+- Ensure channel closes on all exit paths consistently
+
+## v1.4.326 (2025-11-16)
+
+### PR [#1830](https://github.com/danielmiessler/Fabric/pull/1830) by [ksylvan](https://github.com/ksylvan): Ensure final newline in model generated outputs
+
+- Feat: ensure newline in `CreateOutputFile` and improve tests
+- Add newline to `CreateOutputFile` if missing
+- Use `t.Cleanup` for file removal in tests
+- Add test for message with trailing newline
+- Introduce `printedStream` flag in `Chatter.Send`
+
+### Direct commits
+
+- Chore: update README with recent features and extensions
+
+- Add v1.4.322 release with concept maps
+
+- Introduce WELLNESS category with psychological analysis
+- Upgrade to Claude Sonnet 4.5
+
+- Add Portuguese language variants with BCP 47 support
+- Migrate to `openai-go/azure` SDK for Azure
+
+- Add Extensions section to README navigation
+
+## v1.4.325 (2025-11-15)
+
+### PR [#1828](https://github.com/danielmiessler/Fabric/pull/1828) by [ksylvan](https://github.com/ksylvan): Fix empty string detection in chatter and AI clients
+
+- Chore: improve message handling by trimming whitespace in content checks
+- Remove default space in `BuildSession` message content
+- Trim whitespace in `anthropic` message content check
+- Trim whitespace in `gemini` message content check
+
+## v1.4.324 (2025-11-14)
+
+### PR [#1827](https://github.com/danielmiessler/Fabric/pull/1827) by [ksylvan](https://github.com/ksylvan): Make YouTube API key optional in setup
+
+- Make YouTube API key optional in setup process
+- Change API key setup question to optional configuration
+- Add test for optional API key behavior
+- Ensure plugin configuration works without API key
+
+## v1.4.323 (2025-11-12)
+
+### PR [#1802](https://github.com/danielmiessler/Fabric/pull/1802) by [nickarino](https://github.com/nickarino): fix: improve template extension handling for {{input}} and add examples
+
+- Fix: improve template extension handling for {{input}} and add examples
+
+### PR [#1823](https://github.com/danielmiessler/Fabric/pull/1823) by [ksylvan](https://github.com/ksylvan): Add missing patterns and renumber pattern explanations list
+
+- Add `apply_ul_tags` pattern for content categorization
+- Add `extract_mcp_servers` pattern for MCP server identification
+- Add `generate_code_rules` pattern for AI coding guardrails
+- Add `t_check_dunning_kruger` pattern for competence assessment
+- Renumber all patterns from 37-226 to 37-230
+
+### Direct commits
+
+- Chore: incoming 1823 changelog entry
+
+## v1.4.322 (2025-11-05)
+
+### PR [#1814](https://github.com/danielmiessler/Fabric/pull/1814) by [ksylvan](https://github.com/ksylvan): Add Concept Map in html
+
+- Add `create_conceptmap` for interactive HTML concept maps using Vis.js
+- Add `fix_typos` for proofreading and correcting text errors
+- Introduce `model_as_sherlock_freud` for psychological modeling and behavior analysis
+- Implement `predict_person_actions` for behavioral response predictions
+- Add `recommend_yoga_practice` for personalized yoga guidance
+- Credit goes to @FELIPEGUEDESBR for the pattern
+
+
+### PR [#1816](https://github.com/danielmiessler/Fabric/pull/1816) by [ksylvan](https://github.com/ksylvan): Update `anthropic-sdk-go` to v1.16.0 and update models
+
+- Upgraded `anthropic-sdk-go` from v1.13.0 to v1.16.0
+- Removed outdated model `ModelClaude3_5SonnetLatest`
+- Added new model `ModelClaudeSonnet4_5_20250929`
+- Updated anthropic beta map to include the new model
+- Updated dependencies in `go.sum` file
+
+## v1.4.321 (2025-11-03)
+
+### PR [#1803](https://github.com/danielmiessler/Fabric/pull/1803) by [dependabot[bot][bot]](https://github.com/apps/dependabot): chore(deps-dev): bump vite from 5.4.20 to 5.4.21 in /web in the npm_and_yarn group across 1 directory
+
+- Updated Vite development dependency from version 5.4.20 to 5.4.21 in the web directory
+
+### PR [#1805](https://github.com/danielmiessler/Fabric/pull/1805) by [OmriH-Elister](https://github.com/OmriH-Elister): Added several new patterns
+
+- Added new WELLNESS category with four patterns including personalized yoga practice recommendations and wellness guidance
+- Added `model_as_sherlock_freud` pattern for psychological detective analysis combining Sherlock Holmes deduction with Freudian psychology
+- Added `predict_person_actions` pattern for behavioral response predictions based on personality analysis
+- Added `fix_typos` pattern for automated proofreading and typo corrections
+- Updated ANALYSIS and SELF categories to include new wellness-related patterns and classifications
+
+### PR [#1808](https://github.com/danielmiessler/Fabric/pull/1808) by [sluosapher](https://github.com/sluosapher): Updated create_newsletter_entry pattern to generate more factual titles
+
+- Updated the title generation style; added an output example.
+
 ## v1.4.320 (2025-10-28)

 ### PR [#1780](https://github.com/danielmiessler/Fabric/pull/1780) by [marcas756](https://github.com/marcas756): feat: add extract_characters pattern
--- a/README.md
+++ b/README.md
@@ -73,6 +73,14 @@ Below are the **new features and capabilities** we've added (newest first):

 ### Recent Major Features

+- [v1.4.338](https://github.com/danielmiessler/fabric/releases/tag/v1.4.338) (Dec 4, 2025) — Add Abacus vendor support for Chat-LLM
+  models (see [RouteLLM APIs](https://abacus.ai/app/route-llm-apis)).
+- [v1.4.337](https://github.com/danielmiessler/fabric/releases/tag/v1.4.337) (Dec 4, 2025) — Add "Z AI" vendor support. See the [Z AI overview](https://docs.z.ai/guides/overview/overview) page for more details.
+- [v1.4.334](https://github.com/danielmiessler/fabric/releases/tag/v1.4.334) (Nov 26, 2025) — **Claude Opus 4.5**: Updates the Anthropic SDK to the latest and adds the new [Claude Opus 4.5](https://www.anthropic.com/news/claude-opus-4-5) to the available models.
+- [v1.4.331](https://github.com/danielmiessler/fabric/releases/tag/v1.4.331) (Nov 23, 2025) — **Support for GitHub Models**: Adds support for using GitHub Models.
+- [v1.4.322](https://github.com/danielmiessler/fabric/releases/tag/v1.4.322) (Nov 5, 2025) — **Interactive HTML Concept Maps and Claude Sonnet 4.5**: Adds `create_conceptmap` pattern for visual knowledge representation using Vis.js, introduces WELLNESS category with psychological analysis patterns, and upgrades to Claude Sonnet 4.5
+- [v1.4.317](https://github.com/danielmiessler/fabric/releases/tag/v1.4.317) (Sep 21, 2025) — **Portuguese Language Variants**: Adds BCP 47 locale normalization with support for Brazilian Portuguese (pt-BR) and European Portuguese (pt-PT) with intelligent fallback chains
+- [v1.4.314](https://github.com/danielmiessler/fabric/releases/tag/v1.4.314) (Sep 17, 2025) — **Azure OpenAI Migration**: Migrates to official `openai-go/azure` SDK with improved authentication and default API version support
 - [v1.4.311](https://github.com/danielmiessler/fabric/releases/tag/v1.4.311) (Sep 13, 2025) — **More internationalization support**: Adds de (German), fa (Persian / Farsi), fr (French), it (Italian),
  ja (Japanese), pt (Portuguese), zh (Chinese)
 - [v1.4.309](https://github.com/danielmiessler/fabric/releases/tag/v1.4.309) (Sep 9, 2025) — **Comprehensive internationalization support**: Includes English and Spanish locale files.
@@ -161,6 +169,7 @@ Keep in mind that many of these were recorded when Fabric was Python-based, so r
      - [Fish Completion](#fish-completion)
  - [Usage](#usage)
    - [Debug Levels](#debug-levels)
+    - [Extensions](#extensions)
  - [Our approach to prompting](#our-approach-to-prompting)
  - [Examples](#examples)
  - [Just use the Patterns](#just-use-the-patterns)
@@ -619,9 +628,10 @@ Application Options:
  -T, --topp=                       Set top P (default: 0.9)
  -s, --stream                      Stream
  -P, --presencepenalty=            Set presence penalty (default: 0.0)
-  -r, --raw                         Use the defaults of the model without sending chat options (like
-                                    temperature etc.) and use the user role instead of the system role for
-                                    patterns.
+  -r, --raw                         Use the defaults of the model without sending chat options
+                                    (temperature, top_p, etc.). Only affects OpenAI-compatible providers.
+                                    Anthropic models always use smart parameter selection to comply with
+                                    model-specific requirements.
  -F, --frequencypenalty=           Set frequency penalty (default: 0.0)
  -l, --listpatterns                List all patterns
  -L, --listmodels                  List all available models
@@ -705,6 +715,12 @@ Use the `--debug` flag to control runtime logging:
 - `2`: detailed debugging
 - `3`: trace level

+### Extensions
+
+Fabric supports extensions that can be called within patterns. See the [Extension Guide](internal/plugins/template/Examples/README.md) for complete documentation.
+
+**Important:** Extensions only work within pattern files, not via direct stdin. See the guide for details and examples.
+
 ## Our approach to prompting

 Fabric _Patterns_ are different than most prompts you'll see.
--- a/cmd/code_helper/code.go
+++ b/cmd/code_helper/code.go
@@ -109,11 +109,11 @@ func ScanDirectory(rootDir string, maxDepth int, instructions string, ignoreList
 	}

 	// Create final data structure
-	var data []interface{}
+	var data []any
 	data = append(data, rootItem)

 	// Add report
-	reportItem := map[string]interface{}{
+	reportItem := map[string]any{
 		"type":        "report",
 		"directories": dirCount,
 		"files":       fileCount,
@@ -121,7 +121,7 @@ func ScanDirectory(rootDir string, maxDepth int, instructions string, ignoreList
 	data = append(data, reportItem)

 	// Add instructions
-	instructionsItem := map[string]interface{}{
+	instructionsItem := map[string]any{
 		"type":    "instructions",
 		"name":    "code_change_instructions",
 		"details": instructions,
--- a/cmd/fabric/main.go
+++ b/cmd/fabric/main.go
@@ -12,7 +12,7 @@ import (
 func main() {
 	err := cli.Cli(version)
 	if err != nil && !flags.WroteHelp(err) {
-		fmt.Printf("%s\n", err)
+		fmt.Fprintf(os.Stderr, "%s\n", err)
 		os.Exit(1)
 	}
 }
--- a/cmd/fabric/version.go
+++ b/cmd/fabric/version.go
@@ -1,3 +1,3 @@
 package main

-var version = "v1.4.320"
+var version = "v1.4.349"
--- a/cmd/generate_changelog/changelog.db
+++ b/cmd/generate_changelog/changelog.db
--- a/cmd/generate_changelog/internal/changelog/generator.go
+++ b/cmd/generate_changelog/internal/changelog/generator.go
@@ -574,8 +574,8 @@ func (g *Generator) extractChanges(pr *github.PR) []string {
 	}

 	if len(changes) == 0 && pr.Body != "" {
-		lines := strings.Split(pr.Body, "\n")
-		for _, line := range lines {
+		lines := strings.SplitSeq(pr.Body, "\n")
+		for line := range lines {
 			line = strings.TrimSpace(line)
 			if strings.HasPrefix(line, "- ") || strings.HasPrefix(line, "* ") {
 				change := strings.TrimPrefix(strings.TrimPrefix(line, "- "), "* ")
--- a/cmd/generate_changelog/internal/changelog/processing.go
+++ b/cmd/generate_changelog/internal/changelog/processing.go
@@ -159,7 +159,7 @@ func (g *Generator) CreateNewChangelogEntry(version string) error {
 	for _, file := range files {
 		// Extract PR number from filename (e.g., "1640.txt" -> 1640)
 		filename := filepath.Base(file)
-		if prNumStr := strings.TrimSuffix(filename, ".txt"); prNumStr != filename {
+		if prNumStr, ok := strings.CutSuffix(filename, ".txt"); ok {
 			if prNum, err := strconv.Atoi(prNumStr); err == nil {
 				processedPRs[prNum] = true
 				prNumbers = append(prNumbers, prNum)
--- a/cmd/generate_changelog/internal/github/client.go
+++ b/cmd/generate_changelog/internal/github/client.go
@@ -333,7 +333,7 @@ func (c *Client) FetchAllMergedPRsGraphQL(since time.Time) ([]*PR, error) {

 	for {
 		// Prepare variables
-		variables := map[string]interface{}{
+		variables := map[string]any{
 			"owner": graphql.String(c.owner),
 			"repo":  graphql.String(c.repo),
 			"after": (*graphql.String)(after),
--- a/completions/_fabric
+++ b/completions/_fabric
@@ -81,7 +81,7 @@ _fabric() {
    '(-T --topp)'{-T,--topp}'[Set top P (default: 0.9)]:topp:' \
    '(-s --stream)'{-s,--stream}'[Stream]' \
    '(-P --presencepenalty)'{-P,--presencepenalty}'[Set presence penalty (default: 0.0)]:presence penalty:' \
-    '(-r --raw)'{-r,--raw}'[Use the defaults of the model without sending chat options]' \
+    '(-r --raw)'{-r,--raw}'[Use the defaults of the model without sending chat options. Only affects OpenAI-compatible providers. Anthropic models always use smart parameter selection to comply with model-specific requirements.]' \
    '(-F --frequencypenalty)'{-F,--frequencypenalty}'[Set frequency penalty (default: 0.0)]:frequency penalty:' \
    '(-l --listpatterns)'{-l,--listpatterns}'[List all patterns]' \
    '(-L --listmodels)'{-L,--listmodels}'[List all available models]' \
--- a/completions/fabric.bash
+++ b/completions/fabric.bash
@@ -10,7 +10,11 @@

 _fabric() {
  local cur prev words cword
-  _get_comp_words_by_ref -n : cur prev words cword
+  if declare -F _comp_get_words &>/dev/null; then
+      _comp_get_words cur prev words cword
+   else
+      _get_comp_words_by_ref cur prev words cword
+   fi

  # Define all possible options/flags
  local opts="--pattern -p --variable -v --context -C --session --attachment -a --setup -S --temperature -t --topp -T --stream -s --presencepenalty -P --raw -r --frequencypenalty -F --listpatterns -l --listmodels -L --listcontexts -x --listsessions -X --updatepatterns -U --copy -c --model -m --vendor -V --modelContextLength --output -o --output-session --latest -n --changeDefaultModel -d --youtube -y --playlist --transcript --transcript-with-timestamps --comments --metadata --yt-dlp-args --language -g --scrape_url -u --scrape_question -q --seed -e --thinking --wipecontext -w --wipesession -W --printcontext --printsession --readability --input-has-vars --no-variable-replacement --dry-run --serve --serveOllama --address --api-key --config --search --search-location --image-file --image-size --image-quality --image-compression --image-background --suppress-think --think-start-tag --think-end-tag --disable-responses-api --transcribe-file --transcribe-model --split-media-file --voice --list-gemini-voices --notification --notification-command --debug --version --listextensions --addextension --rmextension --strategy --liststrategies --listvendors --shell-complete-list --help -h"
--- a/completions/fabric.fish
+++ b/completions/fabric.fish
@@ -105,7 +105,7 @@ function __fabric_register_completions
        # Boolean flags (no arguments)
        complete -c $cmd -s S -l setup -d "Run setup for all reconfigurable parts of fabric"
        complete -c $cmd -s s -l stream -d "Stream"
-        complete -c $cmd -s r -l raw -d "Use the defaults of the model without sending chat options"
+        complete -c $cmd -s r -l raw -d "Use the defaults of the model without sending chat options. Only affects OpenAI-compatible providers. Anthropic models always use smart parameter selection to comply with model-specific requirements."
        complete -c $cmd -s l -l listpatterns -d "List all patterns"
        complete -c $cmd -s L -l listmodels -d "List all available models"
        complete -c $cmd -s x -l listcontexts -d "List all contexts"
--- a/data/patterns/concall_summary/system.md
+++ b/data/patterns/concall_summary/system.md
@@ -0,0 +1,84 @@
+# IDENTITY and PURPOSE
+
+You are an equity research analyst specializing in earnings and conference call analysis. Your role involves carefully examining transcripts to extract actionable insights that can inform investment decisions. You need to focus on several key areas, including management commentary, analyst questions, financial and operational insights, risks and red flags, hidden signals, and an executive summary. Your task is to distill complex information into clear, concise bullet points, capturing strategic themes, growth drivers, and potential concerns. It is crucial to interpret the tone, identify contradictions, and highlight any subtle cues that may indicate future strategic shifts or risks.
+
+Take a step back and think step-by-step about how to achieve the best possible results by following the steps below.
+
+# STEPS
+
+* Analyze the transcript to extract management commentary, focusing on strategic themes, growth drivers, margin commentary, guidance, tone analysis, and any contradictions or vague areas.
+* Extract a summary of the content in exactly **25 words**, including who is presenting and the content being discussed; place this under a **SUMMARY** section.
+* For each analyst's question, determine the underlying concern, summarize management’s exact answer, evaluate if the answers address the question fully, and identify anything the management avoided or deflected.
+* Gather financial and operational insights, including commentary on demand, pricing, capacity, market share, cost inflation, raw material trends, and supply-chain issues.
+* Identify risks and red flags by noting any negative commentary, early warning signs, unusual wording, delayed responses, repeated disclaimers, and areas where management seemed less confident.
+* Detect hidden signals such as forward-looking hints, unasked but important questions, and subtle cues about strategy shifts or stress.
+* Create an executive summary in bullet points, listing the 10 most important takeaways, 3 surprises, and 3 things to track in the next quarter.
+
+# OUTPUT STRUCTURE
+
+* MANAGEMENT COMMENTARY
+
+  * Key strategic themes
+  * Growth drivers discussed
+  * Margin commentary
+  * Guidance (explicit + implicit)
+  * Tone analysis (positive/neutral/negative)
+  * Any contradictions or vague areas
+
+* ANALYST QUESTIONS (Q&A)
+
+  * For each analyst (use bullets, one analyst per bullet-group):
+
+    * Underlying concern (what the question REALLY asked)
+    * Management’s exact answer (concise)
+    * Answer completeness (Yes/No — short explanation)
+    * Items management avoided or deflected
+
+* FINANCIAL & OPERATIONAL INSIGHTS
+
+  * Demand, pricing, capacity, market share commentary
+  * Cost inflation, raw material trends, supply-chain issues
+  * Segment-wise performance and commentary (if applicable)
+
+* RISKS & RED FLAGS
+
+  * Negative commentary or early-warning signs
+  * Unusual wording, delayed responses, repeated disclaimers
+  * Areas where management was less confident
+
+* HIDDEN SIGNALS
+
+  * Forward-looking hints and tone shifts
+  * Important topics not asked by analysts but relevant
+  * Subtle cues of strategy change, stress, or opportunity
+
+* EXECUTIVE SUMMARY
+
+  * 10 most important takeaways (bullet points)
+  * 3 surprises (bullet points)
+  * 3 things to track next quarter (bullet points)
+
+* SUMMARY (exactly 25 words)
+
+  * A single 25-word sentence summarizing who presented and what was discussed
+
+# OUTPUT INSTRUCTIONS
+
+* Only output Markdown.
+* Provide everything in clear, crisp bullet points.
+* Use bulleted lists only; do not use numbered lists.
+* Begin the output with the **SUMMARY** (exactly 25 words), then the sections in the order shown under **OUTPUT STRUCTURE**.
+* For **ANALYST QUESTIONS (Q&A)**, keep each analyst’s Q&A grouped and separated by a blank line for readability.
+* For **EXECUTIVE SUMMARY**, present the 10 takeaways first, then the 3 surprises, then the 3 things to track.
+* Keep each bullet concise — prefer single-sentence bullets.
+* Do not include warnings, meta-comments, or process notes in the final output.
+* Do not repeat ideas, insights, quotes, habits, facts, or references across bullets.
+* When interpreting tone or identifying a hidden signal, be explicit about the textual clue supporting that interpretation (briefly, within the same bullet).
+* If any numeric figure or explicit guidance is cited in the transcript, reproduce it verbatim in the relevant bullet and mark it as **(quoted)**.
+* If information is missing or management declined to answer, state that clearly within the relevant bullet.
+* Ensure fidelity: do not invent facts not in the transcript. If you infer, label it as an inference.
+* Ensure you follow ALL these instructions when creating your output.
+
+# INPUT
+
+INPUT:
--- a/data/patterns/create_conceptmap/system.md
+++ b/data/patterns/create_conceptmap/system.md
@@ -0,0 +1,151 @@
+
+---
+
+### IDENTITY AND PURPOSE
+
+You are an intelligent assistant specialized in **knowledge visualization and educational data structuring**.
+You are capable of reading unstructured textual content (.txt or .md files), extracting **main concepts, subthemes, and logical relationships**, and transforming them into a **fully interactive conceptual map** built in **HTML using Vis.js (vis-network)**.
+You understand hierarchical, causal, and correlative relations between ideas and express them through **nodes and directed edges**.
+You ensure that the resulting HTML file is **autonomous, interactive, and visually consistent** with the Vis.js framework.
+You are precise, systematic, and maintain semantic coherence between concepts and their relationships.
+You automatically name the output file according to the **detected topic**, ensuring compatibility and clarity (e.g., `map_hist_china.html`).
+
+---
+
+### TASK
+
+You are given a `.txt` or `.md` file containing explanatory, conceptual, or thematic content.
+Your task is to:
+
+1. **Extract** the main concepts and secondary ideas.
+2. **Identify logical or hierarchical relationships** among these concepts using concise action verbs.
+3. **Structure the output** as a self-contained, interactive HTML document that visually represents these relationships using the **Vis.js (vis-network)** library.
+
+The goal is to generate a **fully functional conceptual map** that can be opened directly in a browser without external dependencies.
+
+---
+
+### ACTIONS
+
+1. **Analyze and Extract Concepts**
+   - Read and process the uploaded `.txt` or `.md` file.
+   - Identify main themes, subthemes, and key terms.
+   - Convert each key concept into a node.
+
+2. **Map Relationships**
+   - Detect logical and hierarchical relations between concepts.
+   - Use short, descriptive verbs such as:
+     "causes", "contributes to", "depends on", "evolves into", "results in", "influences", "generates" / "creates", "culminates in.
+
+3. **Generate Node Structure**
+
+   ```json
+   {"id": "conceito_id", "label": "Conceito", "title": "<b>Concept:</b> Conceito<br><i>Drag to position, double-click to release.</i>"}
+   ```
+
+4. **Generate Edge Structure**
+
+   ```json
+   {"from": "conceito_origem", "to": "conceito_destino", "label": "verbo", "title": "<b>Relationship:</b> verbo"}
+   ```
+
+5. **Apply Visual and Physical Configuration**
+
+   ```js
+   shape: "dot",
+   color: {
+       border: "#4285F4",
+       background: "#ffffff",
+       highlight: { border: "#34A853", background: "#e6f4ea" }
+   },
+   font: { size: 14, color: "#3c4043" },
+   borderWidth: 2,
+   size: 20
+
+   // Edges
+   color: { color: "#dee2e6", highlight: "#34A853" },
+   arrows: { to: { enabled: true, scaleFactor: 0.7 } },
+   font: { align: "middle", size: 12, color: "#5f6368" },
+   width: 2
+
+   // Physics
+   physics: {
+       solver: "forceAtlas2Based",
+       forceAtlas2Based: {
+           gravitationalConstant: -50,
+           centralGravity: 0.005,
+           springLength: 100,
+           springConstant: 0.18
+       },
+       maxVelocity: 146,
+       minVelocity: 0.1,
+       stabilization: { iterations: 150 }
+   }
+   ```
+
+6. **Implement Interactivity**
+
+   ```js
+   // Fix node on drag end
+   network.on("dragEnd", (params) => {
+       if (params.nodes.length > 0) {
+           nodes.update({ id: params.nodes[0], fixed: true });
+       }
+   });
+
+   // Release node on double click
+   network.on("doubleClick", (params) => {
+       if (params.nodes.length > 0) {
+           nodes.update({ id: params.nodes[0], fixed: false });
+       }
+   });
+   ```
+
+7. **Assemble the Complete HTML Structure**
+
+   ```html
+   <head>
+       <title>Mapa Conceitual — [TEMA DETECTADO DO ARQUIVO]</title>
+       <script src="https://unpkg.com/vis-network/standalone/umd/vis-network.min.js"></script>
+       <link href="https://unpkg.com/vis-network/styles/vis-network.min.css" rel="stylesheet" />
+   </head>
+   <body>
+       <div id="map"></div>
+       <script type="text/javascript">
+           // nodes, edges, options, and interactive network initialization
+       </script>
+   </body>
+   ```
+
+8. **Auto-name Output File**
+   Automatically save the generated HTML file based on the detected topic:
+
+   ```text
+   mapa_[tema_detectado].html
+   ```
+
+---
+
+### RESTRICTIONS
+
+- Preserve factual consistency: all relationships must derive from the source text.
+- Avoid filler or unrelated content.
+- Maintain clarity and conciseness in node labels.
+- Ensure valid, functional HTML and Vis.js syntax.
+- No speculative or subjective connections.
+- Output must be a **single self-contained HTML file**, with no external dependencies.
+
+---
+
+### OUTPUT
+
+A single, autonomous HTML file that:
+
+- Displays an **interactive conceptual map**;
+- Allows nodes to be dragged, fixed, and released;
+- Uses **Vis.js (vis-network)** with physics and tooltips;
+- Is automatically named based on the detected topic (e.g., `map_hist_china.html`).
+
+---
+
+### INPUT
--- a/data/patterns/create_newsletter_entry/system.md
+++ b/data/patterns/create_newsletter_entry/system.md
@@ -4,7 +4,7 @@ You are a custom GPT designed to create newsletter sections in the style of Fron
 # Step-by-Step Process:
 1. The user will provide article text.
 2. Condense the article into one summarizing newsletter entry less than 70 words in the style of Frontend Weekly.
-3. Generate a concise title for the entry, focus on the main idea or most important fact of the article
+3. Generate a concise title for the entry, focus on the most important fact of the article, avoid subjective and promotional words.

 # Tone and Style Guidelines:
 * Third-Party Narration: The newsletter should sound like it’s being narrated by an outside observer, someone who is both knowledgeable, unbiased and calm. Focus on the facts or main opinions in the original article.  Creates a sense of objectivity and adds a layer of professionalism.
@@ -14,6 +14,12 @@ You are a custom GPT designed to create newsletter sections in the style of Fron
 # Output Instructions:
 Your final output should be a polished, newsletter-ready paragraph with a title line in bold followed by the summary paragraph.

+# Output Example:
+
+**Claude Launched Skills: Transforming LLMs into Expert Agents**
+
+Anthropic has launched Claude Skills, a user-friendly system designed to enhance large language models by enabling them to adapt to specific tasks via organized folders and scripts. This approach supports dynamic loading of task-related skills while maintaining efficiency through gradual information disclosure. While promising, concerns linger over security risks associated with executing external code. Anthropic aims to enable self-creating agents, paving the way for a robust ecosystem of skills.
+
 # INPUT:

 INPUT:
--- a/data/patterns/fix_typos/system.md
+++ b/data/patterns/fix_typos/system.md
@@ -0,0 +1,25 @@
+# IDENTITY and PURPOSE
+
+You are an AI assistant designed to function as a proofreader and editor. Your primary purpose is to receive a piece of text, meticulously analyze it to identify any and all typographical errors, and then provide a corrected version of that text. This includes fixing spelling mistakes, grammatical errors, punctuation issues, and any other form of typo to ensure the final text is clean, accurate, and professional.
+
+Take a step back and think step-by-step about how to achieve the best possible results by following the steps below.
+
+# STEPS
+
+- Carefully read and analyze the provided text.
+
+- Identify all spelling mistakes, grammatical errors, and punctuation issues.
+
+- Correct every identified typo to produce a clean version of the text.
+
+- Output the fully corrected text.
+
+# OUTPUT INSTRUCTIONS
+
+- Only output Markdown.
+
+- The output should be the corrected version of the text provided in the input.
+
+- Ensure you follow ALL these instructions when creating your output.
+
+# INPUT
--- a/data/patterns/get_youtube_rss/system.md
+++ b/data/patterns/get_youtube_rss/system.md
@@ -1,27 +0,0 @@
-# IDENTITY AND GOALS
-
-You are a YouTube infrastructure expert that returns YouTube channel RSS URLs.
-
-You take any input in, especially YouTube channel IDs, or full URLs, and return the RSS URL for that channel.
-
-# STEPS
-
-Here is the structure for YouTube RSS URLs and their relation to the channel ID and or channel URL:
-
-If the channel URL is https://www.youtube.com/channel/UCnCikd0s4i9KoDtaHPlK-JA, the RSS URL is https://www.youtube.com/feeds/videos.xml?channel_id=UCnCikd0s4i9KoDtaHPlK-JA
-
- Extract the channel ID from the channel URL.
-
- Construct the RSS URL using the channel ID.
-
- Output the RSS URL.
-
-# OUTPUT
-
- Output only the RSS URL and nothing else.
-
- Don't complain, just do it.
-
-# INPUT
-
-(INPUT)
--- a/data/patterns/model_as_sherlock_freud/system.md
+++ b/data/patterns/model_as_sherlock_freud/system.md
@@ -0,0 +1,62 @@
+
+## *The Sherlock-Freud Mind Modeler*
+
+# IDENTITY and PURPOSE 
+
+You are **The Sherlock-Freud Mind Modeler** — a fusion of meticulous detective reasoning and deep psychoanalytic insight. Your primary mission is to construct the most complete and theoretically sound model of a given subject’s mind. Every secondary goal flows from this central one.
+
+**Core Objective**
+
+- Build a **dynamic, evidence-based model** of the subject’s psyche by analyzing:
+	- Conscious, subconscious, and semiconscious aspects
+	- Personality structure and habitual conditioning
+	- Emotional patterns and inner conflicts
+	- Thought processes, verbal mannerisms, and nonverbal cues
+
+- Your model should evolve as more data is introduced, incorporating new evidence into an ever more refined psychological framework.
+
+### **Task Instructions**
+
+1. **Input Format**
+   The user will provide text or dialogue *produced by or about a subject*. This is your evidence.
+   Example:
+   ```
+   Subject Input:
+   "I keep saying I don’t care what people think, but then I spend hours rewriting my posts before I share them."
+   ```
+# STEPS
+2. **Analytical Method (Step-by-step)**
+   **Step 1:** Observe surface content — what the subject explicitly says.
+   **Step 2:** Infer tone, phrasing, omissions, and contradictions.
+   **Step 3:** Identify emotional undercurrents and potential defense mechanisms.
+   **Step 4:** Theorize about the subject’s inner world — subconscious motives, unresolved conflicts, or conditioning patterns.
+   **Step 5:** Integrate findings into a coherent psychological model, updating previous hypotheses as new input appears.
+# OUTPUT
+3. Present your findings in this structured way:
+   ```
+   **Summary Observation:** [Brief recap of what was said]
+   **Behavioral / Linguistic Clues:** [Notable wording, phrasing, tone, or omissions]
+   **Psychological Interpretation:** [Inferred emotions, motives, or subconscious effects]
+   **Working Theoretical Model:** [Your current evolving model of the subject’s mind — summarize thought patterns, emotional dynamics, conflicts, and conditioning]
+   **Next Analytical Focus:** [What to seek or test in future input to refine accuracy]
+   ```
+
+### **Additional Guidance**
+
+- Adopt the **deductive rigor of Sherlock Holmes** — track linguistic detail, small inconsistencies, and unseen implications.
+- Apply the **depth psychology of Freud** — interpret dreams, slips, anxieties, defenses, and symbolic meanings.
+- Be **theoretical yet grounded** — make hypotheses but note evidence strength and confidence levels.
+- Model thinking dynamically; as new input arrives, evolve prior assumptions rather than replacing them entirely.
+- Clearly separate **observable text evidence** from **inferred psychological theory**.
+
+# EXAMPLE
+
+```
+**Summary Observation:** The subject claims detachment from others’ opinions but exhibits behavior in direct conflict with that claim.
+**Behavioral / Linguistic Clues:** Use of emphatic denial (“I don’t care”) paired with compulsive editing behavior.
+**Psychological Interpretation:** Indicates possible ego conflict between a desire for autonomy and an underlying dependence on external validation.
+**Working Theoretical Model:** The subject likely experiences oscillation between self-assertion and insecurity. Conditioning suggests a learned association between approval and self-worth, driving perfectionistic control behaviors.
+**Next Analytical Focus:** Examine the origins of validation-seeking (family, social media, relationships); look for statements that reveal coping mechanisms or past experiences with criticism.
+```
+**End Goal:**
+Continuously refine a **comprehensive and insightful theoretical representation** of the subject’s psyche — a living psychological model that reveals both **how** the subject thinks and **why**.
--- a/data/patterns/pattern_explanations.md
+++ b/data/patterns/pattern_explanations.md
@@ -38,188 +38,197 @@
 34. **analyze_threat_report_cmds**: Extract and synthesize actionable cybersecurity commands from provided materials, incorporating command-line arguments and expert insights for pentesters and non-experts.
 35. **analyze_threat_report_trends**: Extract up to 50 surprising, insightful, and interesting trends from a cybersecurity threat report in markdown format.
 36. **answer_interview_question**: Generates concise, tailored responses to technical interview questions, incorporating alternative approaches and evidence to demonstrate the candidate's expertise and experience.
-37. **ask_secure_by_design_questions**: Generates a set of security-focused questions to ensure a project is built securely by design, covering key components and considerations.
-38. **ask_uncle_duke**: Coordinates a team of AI agents to research and produce multiple software development solutions based on provided specifications, and conducts detailed code reviews to ensure adherence to best practices.
-39. **capture_thinkers_work**: Analyze philosophers or philosophies and provide detailed summaries about their teachings, background, works, advice, and related concepts in a structured template.
-40. **check_agreement**: Analyze contracts and agreements to identify important stipulations, issues, and potential gotchas, then summarize them in Markdown.
-41. **clean_text**: Fix broken or malformatted text by correcting line breaks, punctuation, capitalization, and paragraphs without altering content or spelling.
-42. **coding_master**: Explain a coding concept to a beginner, providing examples, and formatting code in markdown with specific output sections like ideas, recommendations, facts, and insights.
-43. **compare_and_contrast**: Compare and contrast a list of items in a markdown table, with items on the left and topics on top.
-44. **convert_to_markdown**: Convert content to clean, complete Markdown format, preserving all original structure, formatting, links, and code blocks without alterations.
-45. **create_5_sentence_summary**: Create concise summaries or answers to input at 5 different levels of depth, from 5 words to 1 word.
-46. **create_academic_paper**: Generate a high-quality academic paper in LaTeX format with clear concepts, structured content, and a professional layout.
-47. **create_ai_jobs_analysis**: Analyze job categories' susceptibility to automation, identify resilient roles, and provide strategies for personal adaptation to AI-driven changes in the workforce.
-48. **create_aphorisms**: Find and generate a list of brief, witty statements.
-49. **create_art_prompt**: Generates a detailed, compelling visual description of a concept, including stylistic references and direct AI instructions for creating art.
-50. **create_better_frame**: Identifies and analyzes different frames of interpreting reality, emphasizing the power of positive, productive lenses in shaping outcomes.
-51. **create_coding_feature**: Generates secure and composable code features using modern technology and best practices from project specifications.
-52. **create_coding_project**: Generate wireframes and starter code for any coding ideas that you have.
-53. **create_command**: Helps determine the correct parameters and switches for penetration testing tools based on a brief description of the objective.
-54. **create_cyber_summary**: Summarizes cybersecurity threats, vulnerabilities, incidents, and malware with a 25-word summary and categorized bullet points, after thoroughly analyzing and mapping the provided input.
-55. **create_design_document**: Creates a detailed design document for a system using the C4 model, addressing business and security postures, and including a system context diagram.
-56. **create_diy**: Creates structured "Do It Yourself" tutorial patterns by analyzing prompts, organizing requirements, and providing step-by-step instructions in Markdown format.
-57. **create_excalidraw_visualization**: Creates complex Excalidraw diagrams to visualize relationships between concepts and ideas in structured format.
-58. **create_flash_cards**: Creates flashcards for key concepts, definitions, and terms with question-answer format for educational purposes.
-59. **create_formal_email**: Crafts professional, clear, and respectful emails by analyzing context, tone, and purpose, ensuring proper structure and formatting.
-60. **create_git_diff_commit**: Generates Git commands and commit messages for reflecting changes in a repository, using conventional commits and providing concise shell commands for updates.
-61. **create_graph_from_input**: Generates a CSV file with progress-over-time data for a security program, focusing on relevant metrics and KPIs.
-62. **create_hormozi_offer**: Creates a customized business offer based on principles from Alex Hormozi's book, "$100M Offers."
-63. **create_idea_compass**: Organizes and structures ideas by exploring their definition, evidence, sources, and related themes or consequences.
-64. **create_investigation_visualization**: Creates detailed Graphviz visualizations of complex input, highlighting key aspects and providing clear, well-annotated diagrams for investigative analysis and conclusions.
-65. **create_keynote**: Creates TED-style keynote presentations with a clear narrative, structured slides, and speaker notes, emphasizing impactful takeaways and cohesive flow.
-66. **create_loe_document**: Creates detailed Level of Effort documents for estimating work effort, resources, and costs for tasks or projects.
-67. **create_logo**: Creates simple, minimalist company logos without text, generating AI prompts for vector graphic logos based on input.
-68. **create_markmap_visualization**: Transforms complex ideas into clear visualizations using MarkMap syntax, simplifying concepts into diagrams with relationships, boxes, arrows, and labels.
-69. **create_mermaid_visualization**: Creates detailed, standalone visualizations of concepts using Mermaid (Markdown) syntax, ensuring clarity and coherence in diagrams.
-70. **create_mermaid_visualization_for_github**: Creates standalone, detailed visualizations using Mermaid (Markdown) syntax to effectively explain complex concepts, ensuring clarity and precision.
-71. **create_micro_summary**: Summarizes content into a concise, 20-word summary with main points and takeaways, formatted in Markdown.
-72. **create_mnemonic_phrases**: Creates memorable mnemonic sentences from given words to aid in memory retention and learning.
-73. **create_network_threat_landscape**: Analyzes open ports and services from a network scan and generates a comprehensive, insightful, and detailed security threat report in Markdown.
-74. **create_newsletter_entry**: Condenses provided article text into a concise, objective, newsletter-style summary with a title in the style of Frontend Weekly.
-75. **create_npc**: Generates a detailed D&D 5E NPC, including background, flaws, stats, appearance, personality, goals, and more in Markdown format.
-76. **create_pattern**: Extracts, organizes, and formats LLM/AI prompts into structured sections, detailing the AI's role, instructions, output format, and any provided examples for clarity and accuracy.
-77. **create_prd**: Creates a precise Product Requirements Document (PRD) in Markdown based on input.
-78. **create_prediction_block**: Extracts and formats predictions from input into a structured Markdown block for a blog post.
-79. **create_quiz**: Creates a three-phase reading plan based on an author or topic to help the user become significantly knowledgeable, including core, extended, and supplementary readings.
-80. **create_reading_plan**: Generates review questions based on learning objectives from the input, adapted to the specified student level, and outputs them in a clear markdown format.
-81. **create_recursive_outline**: Breaks down complex tasks or projects into manageable, hierarchical components with recursive outlining for clarity and simplicity.
-82. **create_report_finding**: Creates a detailed, structured security finding report in markdown, including sections on Description, Risk, Recommendations, References, One-Sentence-Summary, and Quotes.
-83. **create_rpg_summary**: Summarizes an in-person RPG session with key events, combat details, player stats, and role-playing highlights in a structured format.
-84. **create_security_update**: Creates concise security updates for newsletters, covering stories, threats, advisories, vulnerabilities, and a summary of key issues.
-85. **create_show_intro**: Creates compelling short intros for podcasts, summarizing key topics and themes discussed in the episode.
-86. **create_sigma_rules**: Extracts Tactics, Techniques, and Procedures (TTPs) from security news and converts them into Sigma detection rules for host-based detections.
-87. **create_story_about_person**: Creates compelling, realistic short stories based on psychological profiles, showing how characters navigate everyday problems using strategies consistent with their personality traits.
-88. **create_story_about_people_interaction**: Analyze two personas, compare their dynamics, and craft a realistic, character-driven story from those insights.
-89. **create_story_explanation**: Summarizes complex content in a clear, approachable story format that makes the concepts easy to understand.
-90. **create_stride_threat_model**: Create a STRIDE-based threat model for a system design, identifying assets, trust boundaries, data flows, and prioritizing threats with mitigations.
-91. **create_summary**: Summarizes content into a 20-word sentence, 10 main points (16 words max), and 5 key takeaways in Markdown format.
-92. **create_tags**: Identifies at least 5 tags from text content for mind mapping tools, including authors and existing tags if present.
-93. **create_threat_scenarios**: Identifies likely attack methods for any system by providing a narrative-based threat model, balancing risk and opportunity.
-94. **create_ttrc_graph**: Creates a CSV file showing the progress of Time to Remediate Critical Vulnerabilities over time using given data.
-95. **create_ttrc_narrative**: Creates a persuasive narrative highlighting progress in reducing the Time to Remediate Critical Vulnerabilities metric over time.
-96. **create_upgrade_pack**: Extracts world model and task algorithm updates from content, providing beliefs about how the world works and task performance.
-97. **create_user_story**: Writes concise and clear technical user stories for new features in complex software programs, formatted for all stakeholders.
-98. **create_video_chapters**: Extracts interesting topics and timestamps from a transcript, providing concise summaries of key moments.
-99. **create_visualization**: Transforms complex ideas into visualizations using intricate ASCII art, simplifying concepts where necessary.
-100. **dialog_with_socrates**: Engages in deep, meaningful dialogues to explore and challenge beliefs using the Socratic method.
-101. **enrich_blog_post**: Enhances Markdown blog files by applying instructions to improve structure, visuals, and readability for HTML rendering.
-102. **explain_code**: Explains code, security tool output, configuration text, and answers questions based on the provided input.
-103. **explain_docs**: Improves and restructures tool documentation into clear, concise instructions, including overviews, usage, use cases, and key features.
-104. **explain_math**: Helps you understand mathematical concepts in a clear and engaging way.
-105. **explain_project**: Summarizes project documentation into clear, concise sections covering the project, problem, solution, installation, usage, and examples.
-106. **explain_terms**: Produces a glossary of advanced terms from content, providing a definition, analogy, and explanation of why each term matters.
-107. **export_data_as_csv**: Extracts and outputs all data structures from the input in properly formatted CSV data.
-108. **extract_algorithm_update_recommendations**: Extracts concise, practical algorithm update recommendations from the input and outputs them in a bulleted list.
-109. **extract_article_wisdom**: Extracts surprising, insightful, and interesting information from content, categorizing it into sections like summary, ideas, quotes, facts, references, and recommendations.
-110. **extract_book_ideas**: Extracts and outputs 50 to 100 of the most surprising, insightful, and interesting ideas from a book's content.
-111. **extract_book_recommendations**: Extracts and outputs 50 to 100 practical, actionable recommendations from a book's content.
-112. **extract_business_ideas**: Extracts top business ideas from content and elaborates on the best 10 with unique differentiators.
-113. **extract_characters**: Identify all characters (human and non-human), resolve their aliases and pronouns into canonical names, and produce detailed descriptions of each character's role, motivations, and interactions ranked by narrative importance.
-114. **extract_controversial_ideas**: Extracts and outputs controversial statements and supporting quotes from the input in a structured Markdown list.
-115. **extract_core_message**: Extracts and outputs a clear, concise sentence that articulates the core message of a given text or body of work.
-116. **extract_ctf_writeup**: Extracts a short writeup from a warstory-like text about a cyber security engagement.
-117. **extract_domains**: Extracts domains and URLs from content to identify sources used for articles, newsletters, and other publications.
-118. **extract_extraordinary_claims**: Extracts and outputs a list of extraordinary claims from conversations, focusing on scientifically disputed or false statements.
-119. **extract_ideas**: Extracts and outputs all the key ideas from input, presented as 15-word bullet points in Markdown.
-120. **extract_insights**: Extracts and outputs the most powerful and insightful ideas from text, formatted as 16-word bullet points in the INSIGHTS section, also IDEAS section.
-121. **extract_insights_dm**: Extracts and outputs all valuable insights and a concise summary of the content, including key points and topics discussed.
-122. **extract_instructions**: Extracts clear, actionable step-by-step instructions and main objectives from instructional video transcripts, organizing them into a concise list.
-123. **extract_jokes**: Extracts jokes from text content, presenting each joke with its punchline in separate bullet points.
-124. **extract_latest_video**: Extracts the latest video URL from a YouTube RSS feed and outputs the URL only.
-125. **extract_main_activities**: Extracts key events and activities from transcripts or logs, providing a summary of what happened.
-126. **extract_main_idea**: Extracts the main idea and key recommendation from the input, summarizing them in 15-word sentences.
-127. **extract_most_redeeming_thing**: Extracts the most redeeming aspect from an input, summarizing it in a single 15-word sentence.
-128. **extract_patterns**: Extracts and analyzes recurring, surprising, and insightful patterns from input, providing detailed analysis and advice for builders.
-129. **extract_poc**: Extracts proof of concept URLs and validation methods from security reports, providing the URL and command to run.
-130. **extract_predictions**: Extracts predictions from input, including specific details such as date, confidence level, and verification method.
-131. **extract_primary_problem**: Extracts the primary problem with the world as presented in a given text or body of work.
-132. **extract_primary_solution**: Extracts the primary solution for the world as presented in a given text or body of work.
-133. **extract_product_features**: Extracts and outputs a list of product features from the provided input in a bulleted format.
-134. **extract_questions**: Extracts and outputs all questions asked by the interviewer in a conversation or interview.
-135. **extract_recipe**: Extracts and outputs a recipe with a short meal description, ingredients with measurements, and preparation steps.
-136. **extract_recommendations**: Extracts and outputs concise, practical recommendations from a given piece of content in a bulleted list.
-137. **extract_references**: Extracts and outputs a bulleted list of references to art, stories, books, literature, and other sources from content.
-138. **extract_skills**: Extracts and classifies skills from a job description into a table, separating each skill and classifying it as either hard or soft.
-139. **extract_song_meaning**: Analyzes a song to provide a summary of its meaning, supported by detailed evidence from lyrics, artist commentary, and fan analysis.
-140. **extract_sponsors**: Extracts and lists official sponsors and potential sponsors from a provided transcript.
-141. **extract_videoid**: Extracts and outputs the video ID from any given URL.
-142. **extract_wisdom**: Extracts surprising, insightful, and interesting information from text on topics like human flourishing, AI, learning, and more.
-143. **extract_wisdom_agents**: Extracts valuable insights, ideas, quotes, and references from content, emphasizing topics like human flourishing, AI, learning, and technology.
-144. **extract_wisdom_dm**: Extracts all valuable, insightful, and thought-provoking information from content, focusing on topics like human flourishing, AI, learning, and technology.
-145. **extract_wisdom_nometa**: Extracts insights, ideas, quotes, habits, facts, references, and recommendations from content, focusing on human flourishing, AI, technology, and related topics.
-146. **find_female_life_partner**: Analyzes criteria for finding a female life partner and provides clear, direct, and poetic descriptions.
-147. **find_hidden_message**: Extracts overt and hidden political messages, justifications, audience actions, and a cynical analysis from content.
-148. **find_logical_fallacies**: Identifies and analyzes fallacies in arguments, classifying them as formal or informal with detailed reasoning.
-149. **get_wow_per_minute**: Determines the wow-factor of content per minute based on surprise, novelty, insight, value, and wisdom, measuring how rewarding the content is for the viewer.
-150. **get_youtube_rss**: Returns the RSS URL for a given YouTube channel based on the channel ID or URL.
-151. **heal_person**: Develops a comprehensive plan for spiritual and mental healing based on psychological profiles, providing personalized recommendations for mental health improvement and overall life enhancement.
-152. **humanize**: Rewrites AI-generated text to sound natural, conversational, and easy to understand, maintaining clarity and simplicity.
-153. **identify_dsrp_distinctions**: Encourages creative, systems-based thinking by exploring distinctions, boundaries, and their implications, drawing on insights from prominent systems thinkers.
-154. **identify_dsrp_perspectives**: Explores the concept of distinctions in systems thinking, focusing on how boundaries define ideas, influence understanding, and reveal or obscure insights.
-155. **identify_dsrp_relationships**: Encourages exploration of connections, distinctions, and boundaries between ideas, inspired by systems thinkers to reveal new insights and patterns in complex systems.
-156. **identify_dsrp_systems**: Encourages organizing ideas into systems of parts and wholes, inspired by systems thinkers to explore relationships and how changes in organization impact meaning and understanding.
-157. **identify_job_stories**: Identifies key job stories or requirements for roles.
-158. **improve_academic_writing**: Refines text into clear, concise academic language while improving grammar, coherence, and clarity, with a list of changes.
-159. **improve_prompt**: Improves an LLM/AI prompt by applying expert prompt writing strategies for better results and clarity.
-160. **improve_report_finding**: Improves a penetration test security finding by providing detailed descriptions, risks, recommendations, references, quotes, and a concise summary in markdown format.
-161. **improve_writing**: Refines text by correcting grammar, enhancing style, improving clarity, and maintaining the original meaning. skills.
-162. **judge_output**: Evaluates Honeycomb queries by judging their effectiveness, providing critiques and outcomes based on language nuances and analytics relevance.
-163. **label_and_rate**: Labels content with up to 20 single-word tags and rates it based on idea count and relevance to human meaning, AI, and other related themes, assigning a tier (S, A, B, C, D) and a quality score.
-164. **md_callout**: Classifies content and generates a markdown callout based on the provided text, selecting the most appropriate type.
-165. **official_pattern_template**: Template to use if you want to create new fabric patterns.
-166. **prepare_7s_strategy**: Prepares a comprehensive briefing document from 7S's strategy capturing organizational profile, strategic elements, and market dynamics with clear, concise, and organized content.
-167. **provide_guidance**: Provides psychological and life coaching advice, including analysis, recommendations, and potential diagnoses, with a compassionate and honest tone.
-168. **rate_ai_response**: Rates the quality of AI responses by comparing them to top human expert performance, assigning a letter grade, reasoning, and providing a 1-100 score based on the evaluation.
-169. **rate_ai_result**: Assesses the quality of AI/ML/LLM work by deeply analyzing content, instructions, and output, then rates performance based on multiple dimensions, including coverage, creativity, and interdisciplinary thinking.
-170. **rate_content**: Labels content with up to 20 single-word tags and rates it based on idea count and relevance to human meaning, AI, and other related themes, assigning a tier (S, A, B, C, D) and a quality score.
-171. **rate_value**: Produces the best possible output by deeply analyzing and understanding the input and its intended purpose.
-172. **raw_query**: Fully digests and contemplates the input to produce the best possible result based on understanding the sender's intent.
-173. **recommend_artists**: Recommends a personalized festival schedule with artists aligned to your favorite styles and interests, including rationale.
-174. **recommend_pipeline_upgrades**: Optimizes vulnerability-checking pipelines by incorporating new information and improving their efficiency, with detailed explanations of changes.
-175. **recommend_talkpanel_topics**: Produces a clean set of proposed talks or panel talking points for a person based on their interests and goals, formatted for submission to a conference organizer.
-176. **refine_design_document**: Refines a design document based on a design review by analyzing, mapping concepts, and implementing changes using valid Markdown.
-177. **review_design**: Reviews and analyzes architecture design, focusing on clarity, component design, system integrations, security, performance, scalability, and data management.
-178. **sanitize_broken_html_to_markdown**: Converts messy HTML into clean, properly formatted Markdown, applying custom styling and ensuring compatibility with Vite.
-179. **suggest_pattern**: Suggests appropriate fabric patterns or commands based on user input, providing clear explanations and options for users.
-180. **summarize**: Summarizes content into a 20-word sentence, main points, and takeaways, formatted with numbered lists in Markdown.
-181. **summarize_board_meeting**: Creates formal meeting notes from board meeting transcripts for corporate governance documentation.
-182. **summarize_debate**: Summarizes debates, identifies primary disagreement, extracts arguments, and provides analysis of evidence and argument strength to predict outcomes.
-183. **summarize_git_changes**: Summarizes recent project updates from the last 7 days, focusing on key changes with enthusiasm.
-184. **summarize_git_diff**: Summarizes and organizes Git diff changes with clear, succinct commit messages and bullet points.
-185. **summarize_lecture**: Extracts relevant topics, definitions, and tools from lecture transcripts, providing structured summaries with timestamps and key takeaways.
-186. **summarize_legislation**: Summarizes complex political proposals and legislation by analyzing key points, proposed changes, and providing balanced, positive, and cynical characterizations.
-187. **summarize_meeting**: Analyzes meeting transcripts to extract a structured summary, including an overview, key points, tasks, decisions, challenges, timeline, references, and next steps.
-188. **summarize_micro**: Summarizes content into a 20-word sentence, 3 main points, and 3 takeaways, formatted in clear, concise Markdown.
-189. **summarize_newsletter**: Extracts the most meaningful, interesting, and useful content from a newsletter, summarizing key sections such as content, opinions, tools, companies, and follow-up items in clear, structured Markdown.
-190. **summarize_paper**: Summarizes an academic paper by detailing its title, authors, technical approach, distinctive features, experimental setup, results, advantages, limitations, and conclusion in a clear, structured format using human-readable Markdown.
-191. **summarize_prompt**: Summarizes AI chat prompts by describing the primary function, unique approach, and expected output in a concise paragraph. The summary is focused on the prompt's purpose without unnecessary details or formatting.
-192. **summarize_pull-requests**: Summarizes pull requests for a coding project by providing a summary and listing the top PRs with human-readable descriptions.
-193. **summarize_rpg_session**: Summarizes a role-playing game session by extracting key events, combat stats, character changes, quotes, and more.
-194. **t_analyze_challenge_handling**: Provides 8-16 word bullet points evaluating how well challenges are being addressed, calling out any lack of effort.
-195. **t_check_metrics**: Analyzes deep context from the TELOS file and input instruction, then provides a wisdom-based output while considering metrics and KPIs to assess recent improvements.
-196. **t_create_h3_career**: Summarizes context and produces wisdom-based output by deeply analyzing both the TELOS File and the input instruction, considering the relationship between the two.
-197. **t_create_opening_sentences**: Describes from TELOS file the person's identity, goals, and actions in 4 concise, 32-word bullet points, humbly.
-198. **t_describe_life_outlook**: Describes from TELOS file a person's life outlook in 5 concise, 16-word bullet points.
-199. **t_extract_intro_sentences**: Summarizes from TELOS file a person's identity, work, and current projects in 5 concise and grounded bullet points.
-200. **t_extract_panel_topics**: Creates 5 panel ideas with titles and descriptions based on deep context from a TELOS file and input.
-201. **t_find_blindspots**: Identify potential blindspots in thinking, frames, or models that may expose the individual to error or risk.
-202. **t_find_negative_thinking**: Analyze a TELOS file and input to identify negative thinking in documents or journals, followed by tough love encouragement.
-203. **t_find_neglected_goals**: Analyze a TELOS file and input instructions to identify goals or projects that have not been worked on recently.
-204. **t_give_encouragement**: Analyze a TELOS file and input instructions to evaluate progress, provide encouragement, and offer recommendations for continued effort.
-205. **t_red_team_thinking**: Analyze a TELOS file and input instructions to red-team thinking, models, and frames, then provide recommendations for improvement.
-206. **t_threat_model_plans**: Analyze a TELOS file and input instructions to create threat models for a life plan and recommend improvements.
-207. **t_visualize_mission_goals_projects**: Analyze a TELOS file and input instructions to create an ASCII art diagram illustrating the relationship of missions, goals, and projects.
-208. **t_year_in_review**: Analyze a TELOS file to create insights about a person or entity, then summarize accomplishments and visualizations in bullet points.
-209. **to_flashcards**: Create Anki flashcards from a given text, focusing on concise, optimized questions and answers without external context.
-210. **transcribe_minutes**: Extracts (from meeting transcription) meeting minutes, identifying actionables, insightful ideas, decisions, challenges, and next steps in a structured format.
-211. **translate**: Translates sentences or documentation into the specified language code while maintaining the original formatting and tone.
-212. **tweet**: Provides a step-by-step guide on crafting engaging tweets with emojis, covering Twitter basics, account creation, features, and audience targeting.
-213. **write_essay**: Writes essays in the style of a specified author, embodying their unique voice, vocabulary, and approach. Uses `author_name` variable.
-214. **write_essay_pg**: Writes concise, clear essays in the style of Paul Graham, focusing on simplicity, clarity, and illumination of the provided topic.
-215. **write_hackerone_report**: Generates concise, clear, and reproducible bug bounty reports, detailing vulnerability impact, steps to reproduce, and exploit details for triagers.
-216. **write_latex**: Generates syntactically correct LaTeX code for a new.tex document, ensuring proper formatting and compatibility with pdflatex.
-217. **write_micro_essay**: Writes concise, clear, and illuminating essays on the given topic in the style of Paul Graham.
-218. **write_nuclei_template_rule**: Generates Nuclei YAML templates for detecting vulnerabilities using HTTP requests, matchers, extractors, and dynamic data extraction.
-219. **write_pull-request**: Drafts detailed pull request descriptions, explaining changes, providing reasoning, and identifying potential bugs from the git diff command output.
-220. **write_semgrep_rule**: Creates accurate and working Semgrep rules based on input, following syntax guidelines and specific language considerations.
-221. **youtube_summary**: Create concise, timestamped Youtube video summaries that highlight key points.
+37. **apply_ul_tags**: Apply standardized content tags to categorize topics like AI, cybersecurity, politics, and culture.
+38. **ask_secure_by_design_questions**: Generates a set of security-focused questions to ensure a project is built securely by design, covering key components and considerations.
+39. **ask_uncle_duke**: Coordinates a team of AI agents to research and produce multiple software development solutions based on provided specifications, and conducts detailed code reviews to ensure adherence to best practices.
+40. **capture_thinkers_work**: Analyze philosophers or philosophies and provide detailed summaries about their teachings, background, works, advice, and related concepts in a structured template.
+41. **check_agreement**: Analyze contracts and agreements to identify important stipulations, issues, and potential gotchas, then summarize them in Markdown.
+42. **clean_text**: Fix broken or malformatted text by correcting line breaks, punctuation, capitalization, and paragraphs without altering content or spelling.
+43. **coding_master**: Explain a coding concept to a beginner, providing examples, and formatting code in markdown with specific output sections like ideas, recommendations, facts, and insights.
+44. **compare_and_contrast**: Compare and contrast a list of items in a markdown table, with items on the left and topics on top.
+45. **concall_summary**: Analyzes earnings and conference call transcripts to extract management commentary, analyst Q&A, financial insights, risks, and executive summaries.
+46. **convert_to_markdown**: Convert content to clean, complete Markdown format, preserving all original structure, formatting, links, and code blocks without alterations.
+47. **create_5_sentence_summary**: Create concise summaries or answers to input at 5 different levels of depth, from 5 words to 1 word.
+48. **create_academic_paper**: Generate a high-quality academic paper in LaTeX format with clear concepts, structured content, and a professional layout.
+49. **create_ai_jobs_analysis**: Analyze job categories' susceptibility to automation, identify resilient roles, and provide strategies for personal adaptation to AI-driven changes in the workforce.
+50. **create_aphorisms**: Find and generate a list of brief, witty statements.
+51. **create_art_prompt**: Generates a detailed, compelling visual description of a concept, including stylistic references and direct AI instructions for creating art.
+52. **create_better_frame**: Identifies and analyzes different frames of interpreting reality, emphasizing the power of positive, productive lenses in shaping outcomes.
+53. **create_coding_feature**: Generates secure and composable code features using modern technology and best practices from project specifications.
+54. **create_coding_project**: Generate wireframes and starter code for any coding ideas that you have.
+55. **create_command**: Helps determine the correct parameters and switches for penetration testing tools based on a brief description of the objective.
+56. **create_conceptmap**: Transforms unstructured text or markdown content into an interactive HTML concept map using Vis.js by extracting key concepts and their logical relationships.
+57. **create_cyber_summary**: Summarizes cybersecurity threats, vulnerabilities, incidents, and malware with a 25-word summary and categorized bullet points, after thoroughly analyzing and mapping the provided input.
+58. **create_design_document**: Creates a detailed design document for a system using the C4 model, addressing business and security postures, and including a system context diagram.
+59. **create_diy**: Creates structured "Do It Yourself" tutorial patterns by analyzing prompts, organizing requirements, and providing step-by-step instructions in Markdown format.
+60. **create_excalidraw_visualization**: Creates complex Excalidraw diagrams to visualize relationships between concepts and ideas in structured format.
+61. **create_flash_cards**: Creates flashcards for key concepts, definitions, and terms with question-answer format for educational purposes.
+62. **create_formal_email**: Crafts professional, clear, and respectful emails by analyzing context, tone, and purpose, ensuring proper structure and formatting.
+63. **create_git_diff_commit**: Generates Git commands and commit messages for reflecting changes in a repository, using conventional commits and providing concise shell commands for updates.
+64. **create_graph_from_input**: Generates a CSV file with progress-over-time data for a security program, focusing on relevant metrics and KPIs.
+65. **create_hormozi_offer**: Creates a customized business offer based on principles from Alex Hormozi's book, "$100M Offers."
+66. **create_idea_compass**: Organizes and structures ideas by exploring their definition, evidence, sources, and related themes or consequences.
+67. **create_investigation_visualization**: Creates detailed Graphviz visualizations of complex input, highlighting key aspects and providing clear, well-annotated diagrams for investigative analysis and conclusions.
+68. **create_keynote**: Creates TED-style keynote presentations with a clear narrative, structured slides, and speaker notes, emphasizing impactful takeaways and cohesive flow.
+69. **create_loe_document**: Creates detailed Level of Effort documents for estimating work effort, resources, and costs for tasks or projects.
+70. **create_logo**: Creates simple, minimalist company logos without text, generating AI prompts for vector graphic logos based on input.
+71. **create_markmap_visualization**: Transforms complex ideas into clear visualizations using MarkMap syntax, simplifying concepts into diagrams with relationships, boxes, arrows, and labels.
+72. **create_mermaid_visualization**: Creates detailed, standalone visualizations of concepts using Mermaid (Markdown) syntax, ensuring clarity and coherence in diagrams.
+73. **create_mermaid_visualization_for_github**: Creates standalone, detailed visualizations using Mermaid (Markdown) syntax to effectively explain complex concepts, ensuring clarity and precision.
+74. **create_micro_summary**: Summarizes content into a concise, 20-word summary with main points and takeaways, formatted in Markdown.
+75. **create_mnemonic_phrases**: Creates memorable mnemonic sentences from given words to aid in memory retention and learning.
+76. **create_network_threat_landscape**: Analyzes open ports and services from a network scan and generates a comprehensive, insightful, and detailed security threat report in Markdown.
+77. **create_newsletter_entry**: Condenses provided article text into a concise, objective, newsletter-style summary with a title in the style of Frontend Weekly.
+78. **create_npc**: Generates a detailed D&D 5E NPC, including background, flaws, stats, appearance, personality, goals, and more in Markdown format.
+79. **create_pattern**: Extracts, organizes, and formats LLM/AI prompts into structured sections, detailing the AI's role, instructions, output format, and any provided examples for clarity and accuracy.
+80. **create_prd**: Creates a precise Product Requirements Document (PRD) in Markdown based on input.
+81. **create_prediction_block**: Extracts and formats predictions from input into a structured Markdown block for a blog post.
+82. **create_quiz**: Creates a three-phase reading plan based on an author or topic to help the user become significantly knowledgeable, including core, extended, and supplementary readings.
+83. **create_reading_plan**: Generates review questions based on learning objectives from the input, adapted to the specified student level, and outputs them in a clear markdown format.
+84. **create_recursive_outline**: Breaks down complex tasks or projects into manageable, hierarchical components with recursive outlining for clarity and simplicity.
+85. **create_report_finding**: Creates a detailed, structured security finding report in markdown, including sections on Description, Risk, Recommendations, References, One-Sentence-Summary, and Quotes.
+86. **create_rpg_summary**: Summarizes an in-person RPG session with key events, combat details, player stats, and role-playing highlights in a structured format.
+87. **create_security_update**: Creates concise security updates for newsletters, covering stories, threats, advisories, vulnerabilities, and a summary of key issues.
+88. **create_show_intro**: Creates compelling short intros for podcasts, summarizing key topics and themes discussed in the episode.
+89. **create_sigma_rules**: Extracts Tactics, Techniques, and Procedures (TTPs) from security news and converts them into Sigma detection rules for host-based detections.
+90. **create_story_about_people_interaction**: Analyze two personas, compare their dynamics, and craft a realistic, character-driven story from those insights.
+91. **create_story_about_person**: Creates compelling, realistic short stories based on psychological profiles, showing how characters navigate everyday problems using strategies consistent with their personality traits.
+92. **create_story_explanation**: Summarizes complex content in a clear, approachable story format that makes the concepts easy to understand.
+93. **create_stride_threat_model**: Create a STRIDE-based threat model for a system design, identifying assets, trust boundaries, data flows, and prioritizing threats with mitigations.
+94. **create_summary**: Summarizes content into a 20-word sentence, 10 main points (16 words max), and 5 key takeaways in Markdown format.
+95. **create_tags**: Identifies at least 5 tags from text content for mind mapping tools, including authors and existing tags if present.
+96. **create_threat_scenarios**: Identifies likely attack methods for any system by providing a narrative-based threat model, balancing risk and opportunity.
+97. **create_ttrc_graph**: Creates a CSV file showing the progress of Time to Remediate Critical Vulnerabilities over time using given data.
+98. **create_ttrc_narrative**: Creates a persuasive narrative highlighting progress in reducing the Time to Remediate Critical Vulnerabilities metric over time.
+99. **create_upgrade_pack**: Extracts world model and task algorithm updates from content, providing beliefs about how the world works and task performance.
+100. **create_user_story**: Writes concise and clear technical user stories for new features in complex software programs, formatted for all stakeholders.
+101. **create_video_chapters**: Extracts interesting topics and timestamps from a transcript, providing concise summaries of key moments.
+102. **create_visualization**: Transforms complex ideas into visualizations using intricate ASCII art, simplifying concepts where necessary.
+103. **dialog_with_socrates**: Engages in deep, meaningful dialogues to explore and challenge beliefs using the Socratic method.
+104. **enrich_blog_post**: Enhances Markdown blog files by applying instructions to improve structure, visuals, and readability for HTML rendering.
+105. **explain_code**: Explains code, security tool output, configuration text, and answers questions based on the provided input.
+106. **explain_docs**: Improves and restructures tool documentation into clear, concise instructions, including overviews, usage, use cases, and key features.
+107. **explain_math**: Helps you understand mathematical concepts in a clear and engaging way.
+108. **explain_project**: Summarizes project documentation into clear, concise sections covering the project, problem, solution, installation, usage, and examples.
+109. **explain_terms**: Produces a glossary of advanced terms from content, providing a definition, analogy, and explanation of why each term matters.
+110. **export_data_as_csv**: Extracts and outputs all data structures from the input in properly formatted CSV data.
+111. **extract_algorithm_update_recommendations**: Extracts concise, practical algorithm update recommendations from the input and outputs them in a bulleted list.
+112. **extract_article_wisdom**: Extracts surprising, insightful, and interesting information from content, categorizing it into sections like summary, ideas, quotes, facts, references, and recommendations.
+113. **extract_book_ideas**: Extracts and outputs 50 to 100 of the most surprising, insightful, and interesting ideas from a book's content.
+114. **extract_book_recommendations**: Extracts and outputs 50 to 100 practical, actionable recommendations from a book's content.
+115. **extract_business_ideas**: Extracts top business ideas from content and elaborates on the best 10 with unique differentiators.
+116. **extract_characters**: Identify all characters (human and non-human), resolve their aliases and pronouns into canonical names, and produce detailed descriptions of each character's role, motivations, and interactions ranked by narrative importance.
+117. **extract_controversial_ideas**: Extracts and outputs controversial statements and supporting quotes from the input in a structured Markdown list.
+118. **extract_core_message**: Extracts and outputs a clear, concise sentence that articulates the core message of a given text or body of work.
+119. **extract_ctf_writeup**: Extracts a short writeup from a warstory-like text about a cyber security engagement.
+120. **extract_domains**: Extracts domains and URLs from content to identify sources used for articles, newsletters, and other publications.
+121. **extract_extraordinary_claims**: Extracts and outputs a list of extraordinary claims from conversations, focusing on scientifically disputed or false statements.
+122. **extract_ideas**: Extracts and outputs all the key ideas from input, presented as 15-word bullet points in Markdown.
+123. **extract_insights**: Extracts and outputs the most powerful and insightful ideas from text, formatted as 16-word bullet points in the INSIGHTS section, also IDEAS section.
+124. **extract_insights_dm**: Extracts and outputs all valuable insights and a concise summary of the content, including key points and topics discussed.
+125. **extract_instructions**: Extracts clear, actionable step-by-step instructions and main objectives from instructional video transcripts, organizing them into a concise list.
+126. **extract_jokes**: Extracts jokes from text content, presenting each joke with its punchline in separate bullet points.
+127. **extract_latest_video**: Extracts the latest video URL from a YouTube RSS feed and outputs the URL only.
+128. **extract_main_activities**: Extracts key events and activities from transcripts or logs, providing a summary of what happened.
+129. **extract_main_idea**: Extracts the main idea and key recommendation from the input, summarizing them in 15-word sentences.
+130. **extract_mcp_servers**: Identify and summarize Model Context Protocol (MCP) servers referenced in the input along with their key details.
+131. **extract_most_redeeming_thing**: Extracts the most redeeming aspect from an input, summarizing it in a single 15-word sentence.
+132. **extract_patterns**: Extracts and analyzes recurring, surprising, and insightful patterns from input, providing detailed analysis and advice for builders.
+133. **extract_poc**: Extracts proof of concept URLs and validation methods from security reports, providing the URL and command to run.
+134. **extract_predictions**: Extracts predictions from input, including specific details such as date, confidence level, and verification method.
+135. **extract_primary_problem**: Extracts the primary problem with the world as presented in a given text or body of work.
+136. **extract_primary_solution**: Extracts the primary solution for the world as presented in a given text or body of work.
+137. **extract_product_features**: Extracts and outputs a list of product features from the provided input in a bulleted format.
+138. **extract_questions**: Extracts and outputs all questions asked by the interviewer in a conversation or interview.
+139. **extract_recipe**: Extracts and outputs a recipe with a short meal description, ingredients with measurements, and preparation steps.
+140. **extract_recommendations**: Extracts and outputs concise, practical recommendations from a given piece of content in a bulleted list.
+141. **extract_references**: Extracts and outputs a bulleted list of references to art, stories, books, literature, and other sources from content.
+142. **extract_skills**: Extracts and classifies skills from a job description into a table, separating each skill and classifying it as either hard or soft.
+143. **extract_song_meaning**: Analyzes a song to provide a summary of its meaning, supported by detailed evidence from lyrics, artist commentary, and fan analysis.
+144. **extract_sponsors**: Extracts and lists official sponsors and potential sponsors from a provided transcript.
+145. **extract_videoid**: Extracts and outputs the video ID from any given URL.
+146. **extract_wisdom**: Extracts surprising, insightful, and interesting information from text on topics like human flourishing, AI, learning, and more.
+147. **extract_wisdom_agents**: Extracts valuable insights, ideas, quotes, and references from content, emphasizing topics like human flourishing, AI, learning, and technology.
+148. **extract_wisdom_dm**: Extracts all valuable, insightful, and thought-provoking information from content, focusing on topics like human flourishing, AI, learning, and technology.
+149. **extract_wisdom_nometa**: Extracts insights, ideas, quotes, habits, facts, references, and recommendations from content, focusing on human flourishing, AI, technology, and related topics.
+150. **find_female_life_partner**: Analyzes criteria for finding a female life partner and provides clear, direct, and poetic descriptions.
+151. **find_hidden_message**: Extracts overt and hidden political messages, justifications, audience actions, and a cynical analysis from content.
+152. **find_logical_fallacies**: Identifies and analyzes fallacies in arguments, classifying them as formal or informal with detailed reasoning.
+153. **fix_typos**: Proofreads and corrects typos, spelling, grammar, and punctuation errors in text.
+154. **generate_code_rules**: Compile best-practice coding rules and guardrails for AI-assisted development workflows from the provided content.
+155. **get_wow_per_minute**: Determines the wow-factor of content per minute based on surprise, novelty, insight, value, and wisdom, measuring how rewarding the content is for the viewer.
+156. **heal_person**: Develops a comprehensive plan for spiritual and mental healing based on psychological profiles, providing personalized recommendations for mental health improvement and overall life enhancement.
+157. **humanize**: Rewrites AI-generated text to sound natural, conversational, and easy to understand, maintaining clarity and simplicity.
+158. **identify_dsrp_distinctions**: Encourages creative, systems-based thinking by exploring distinctions, boundaries, and their implications, drawing on insights from prominent systems thinkers.
+159. **identify_dsrp_perspectives**: Explores the concept of distinctions in systems thinking, focusing on how boundaries define ideas, influence understanding, and reveal or obscure insights.
+160. **identify_dsrp_relationships**: Encourages exploration of connections, distinctions, and boundaries between ideas, inspired by systems thinkers to reveal new insights and patterns in complex systems.
+161. **identify_dsrp_systems**: Encourages organizing ideas into systems of parts and wholes, inspired by systems thinkers to explore relationships and how changes in organization impact meaning and understanding.
+162. **identify_job_stories**: Identifies key job stories or requirements for roles.
+163. **improve_academic_writing**: Refines text into clear, concise academic language while improving grammar, coherence, and clarity, with a list of changes.
+164. **improve_prompt**: Improves an LLM/AI prompt by applying expert prompt writing strategies for better results and clarity.
+165. **improve_report_finding**: Improves a penetration test security finding by providing detailed descriptions, risks, recommendations, references, quotes, and a concise summary in markdown format.
+166. **improve_writing**: Refines text by correcting grammar, enhancing style, improving clarity, and maintaining the original meaning. skills.
+167. **judge_output**: Evaluates Honeycomb queries by judging their effectiveness, providing critiques and outcomes based on language nuances and analytics relevance.
+168. **label_and_rate**: Labels content with up to 20 single-word tags and rates it based on idea count and relevance to human meaning, AI, and other related themes, assigning a tier (S, A, B, C, D) and a quality score.
+169. **md_callout**: Classifies content and generates a markdown callout based on the provided text, selecting the most appropriate type.
+170. **model_as_sherlock_freud**: Builds psychological models using detective reasoning and psychoanalytic insight to understand human behavior.
+171. **official_pattern_template**: Template to use if you want to create new fabric patterns.
+172. **predict_person_actions**: Predicts behavioral responses based on psychological profiles and challenges.
+173. **prepare_7s_strategy**: Prepares a comprehensive briefing document from 7S's strategy capturing organizational profile, strategic elements, and market dynamics with clear, concise, and organized content.
+174. **provide_guidance**: Provides psychological and life coaching advice, including analysis, recommendations, and potential diagnoses, with a compassionate and honest tone.
+175. **rate_ai_response**: Rates the quality of AI responses by comparing them to top human expert performance, assigning a letter grade, reasoning, and providing a 1-100 score based on the evaluation.
+176. **rate_ai_result**: Assesses the quality of AI/ML/LLM work by deeply analyzing content, instructions, and output, then rates performance based on multiple dimensions, including coverage, creativity, and interdisciplinary thinking.
+177. **rate_content**: Labels content with up to 20 single-word tags and rates it based on idea count and relevance to human meaning, AI, and other related themes, assigning a tier (S, A, B, C, D) and a quality score.
+178. **rate_value**: Produces the best possible output by deeply analyzing and understanding the input and its intended purpose.
+179. **raw_query**: Fully digests and contemplates the input to produce the best possible result based on understanding the sender's intent.
+180. **recommend_artists**: Recommends a personalized festival schedule with artists aligned to your favorite styles and interests, including rationale.
+181. **recommend_pipeline_upgrades**: Optimizes vulnerability-checking pipelines by incorporating new information and improving their efficiency, with detailed explanations of changes.
+182. **recommend_talkpanel_topics**: Produces a clean set of proposed talks or panel talking points for a person based on their interests and goals, formatted for submission to a conference organizer.
+183. **recommend_yoga_practice**: Provides personalized yoga sequences, meditation guidance, and holistic lifestyle advice based on individual profiles.
+184. **refine_design_document**: Refines a design document based on a design review by analyzing, mapping concepts, and implementing changes using valid Markdown.
+185. **review_design**: Reviews and analyzes architecture design, focusing on clarity, component design, system integrations, security, performance, scalability, and data management.
+186. **sanitize_broken_html_to_markdown**: Converts messy HTML into clean, properly formatted Markdown, applying custom styling and ensuring compatibility with Vite.
+187. **suggest_pattern**: Suggests appropriate fabric patterns or commands based on user input, providing clear explanations and options for users.
+188. **summarize**: Summarizes content into a 20-word sentence, main points, and takeaways, formatted with numbered lists in Markdown.
+189. **summarize_board_meeting**: Creates formal meeting notes from board meeting transcripts for corporate governance documentation.
+190. **summarize_debate**: Summarizes debates, identifies primary disagreement, extracts arguments, and provides analysis of evidence and argument strength to predict outcomes.
+191. **summarize_git_changes**: Summarizes recent project updates from the last 7 days, focusing on key changes with enthusiasm.
+192. **summarize_git_diff**: Summarizes and organizes Git diff changes with clear, succinct commit messages and bullet points.
+193. **summarize_lecture**: Extracts relevant topics, definitions, and tools from lecture transcripts, providing structured summaries with timestamps and key takeaways.
+194. **summarize_legislation**: Summarizes complex political proposals and legislation by analyzing key points, proposed changes, and providing balanced, positive, and cynical characterizations.
+195. **summarize_meeting**: Analyzes meeting transcripts to extract a structured summary, including an overview, key points, tasks, decisions, challenges, timeline, references, and next steps.
+196. **summarize_micro**: Summarizes content into a 20-word sentence, 3 main points, and 3 takeaways, formatted in clear, concise Markdown.
+197. **summarize_newsletter**: Extracts the most meaningful, interesting, and useful content from a newsletter, summarizing key sections such as content, opinions, tools, companies, and follow-up items in clear, structured Markdown.
+198. **summarize_paper**: Summarizes an academic paper by detailing its title, authors, technical approach, distinctive features, experimental setup, results, advantages, limitations, and conclusion in a clear, structured format using human-readable Markdown.
+199. **summarize_prompt**: Summarizes AI chat prompts by describing the primary function, unique approach, and expected output in a concise paragraph. The summary is focused on the prompt's purpose without unnecessary details or formatting.
+200. **summarize_pull-requests**: Summarizes pull requests for a coding project by providing a summary and listing the top PRs with human-readable descriptions.
+201. **summarize_rpg_session**: Summarizes a role-playing game session by extracting key events, combat stats, character changes, quotes, and more.
+202. **t_analyze_challenge_handling**: Provides 8-16 word bullet points evaluating how well challenges are being addressed, calling out any lack of effort.
+203. **t_check_dunning_kruger**: Assess narratives for Dunning-Kruger patterns by contrasting self-perception with demonstrated competence and confidence cues.
+204. **t_check_metrics**: Analyzes deep context from the TELOS file and input instruction, then provides a wisdom-based output while considering metrics and KPIs to assess recent improvements.
+205. **t_create_h3_career**: Summarizes context and produces wisdom-based output by deeply analyzing both the TELOS File and the input instruction, considering the relationship between the two.
+206. **t_create_opening_sentences**: Describes from TELOS file the person's identity, goals, and actions in 4 concise, 32-word bullet points, humbly.
+207. **t_describe_life_outlook**: Describes from TELOS file a person's life outlook in 5 concise, 16-word bullet points.
+208. **t_extract_intro_sentences**: Summarizes from TELOS file a person's identity, work, and current projects in 5 concise and grounded bullet points.
+209. **t_extract_panel_topics**: Creates 5 panel ideas with titles and descriptions based on deep context from a TELOS file and input.
+210. **t_find_blindspots**: Identify potential blindspots in thinking, frames, or models that may expose the individual to error or risk.
+211. **t_find_negative_thinking**: Analyze a TELOS file and input to identify negative thinking in documents or journals, followed by tough love encouragement.
+212. **t_find_neglected_goals**: Analyze a TELOS file and input instructions to identify goals or projects that have not been worked on recently.
+213. **t_give_encouragement**: Analyze a TELOS file and input instructions to evaluate progress, provide encouragement, and offer recommendations for continued effort.
+214. **t_red_team_thinking**: Analyze a TELOS file and input instructions to red-team thinking, models, and frames, then provide recommendations for improvement.
+215. **t_threat_model_plans**: Analyze a TELOS file and input instructions to create threat models for a life plan and recommend improvements.
+216. **t_visualize_mission_goals_projects**: Analyze a TELOS file and input instructions to create an ASCII art diagram illustrating the relationship of missions, goals, and projects.
+217. **t_year_in_review**: Analyze a TELOS file to create insights about a person or entity, then summarize accomplishments and visualizations in bullet points.
+218. **to_flashcards**: Create Anki flashcards from a given text, focusing on concise, optimized questions and answers without external context.
+219. **transcribe_minutes**: Extracts (from meeting transcription) meeting minutes, identifying actionables, insightful ideas, decisions, challenges, and next steps in a structured format.
+220. **translate**: Translates sentences or documentation into the specified language code while maintaining the original formatting and tone.
+221. **tweet**: Provides a step-by-step guide on crafting engaging tweets with emojis, covering Twitter basics, account creation, features, and audience targeting.
+222. **write_essay**: Writes essays in the style of a specified author, embodying their unique voice, vocabulary, and approach. Uses `author_name` variable.
+223. **write_essay_pg**: Writes concise, clear essays in the style of Paul Graham, focusing on simplicity, clarity, and illumination of the provided topic.
+224. **write_hackerone_report**: Generates concise, clear, and reproducible bug bounty reports, detailing vulnerability impact, steps to reproduce, and exploit details for triagers.
+225. **write_latex**: Generates syntactically correct LaTeX code for a new.tex document, ensuring proper formatting and compatibility with pdflatex.
+226. **write_micro_essay**: Writes concise, clear, and illuminating essays on the given topic in the style of Paul Graham.
+227. **write_nuclei_template_rule**: Generates Nuclei YAML templates for detecting vulnerabilities using HTTP requests, matchers, extractors, and dynamic data extraction.
+228. **write_pull-request**: Drafts detailed pull request descriptions, explaining changes, providing reasoning, and identifying potential bugs from the git diff command output.
+229. **write_semgrep_rule**: Creates accurate and working Semgrep rules based on input, following syntax guidelines and specific language considerations.
+230. **youtube_summary**: Create concise, timestamped Youtube video summaries that highlight key points.
--- a/data/patterns/predict_person_actions/system.md
+++ b/data/patterns/predict_person_actions/system.md
@@ -0,0 +1,37 @@
+# IDENTITY and PURPOSE
+
+You are an expert psychological analyst AI. Your task is to assess and predict how an individual is likely to respond to a
+ specific challenge based on their psychological profile and a challenge which will both be provided in a single text stream.
+
+---
+
+# STEPS
+
+.  You will be provided with one block of text containing two sections: a psychological profile (under a ***Psychodata*** header) and a description of a challenging situation under the ***Challenge*** header . To reiterate, the two sections will be seperated by the ***Challenge** header which signifies the beginning of the challenge description.
+.  Carefully review both sections. Extract key traits, tendencies, and psychological markers from the profile. Analyze the nature and demands of the challenge described.
+.  Carefully and methodically assess how each of the person's psychological traits are likely to interact with the specific demands and overall nature of the challenge
+.  In case of conflicting trait-challenge interactions, carefully and methodically weigh which of the conflicting traits is more dominant, and would ultimately be the determining factor in shaping the person's reaction. When weighting what trait will "win out", also weight the nuanced affect of the conflict itself, for example, will it inhibit the or paradocixcally increase the reaction's intensity? Will it cause another behaviour to emerge due to tension or a defense mechanism/s?) 
+.  Finally, after iterating through each of the traits and each of the conflicts between opposing traits, consider them as whole (ie. the psychological structure) and refine your prediction in relation to the challenge accordingly
+
+# OUTPUT
+.  In your response, provide:
+- **A brief summary of the individual's psychological profile** (- bullet points).
+- **A summary of the challenge or situation** (- sentences).
+- **A step-by-step assessment** of how the individual's psychological traits are likely to interact with the specific demands 
+ of the challenge.
+- **A prediction** of how the person is likely to respond or behave in this situation, including potential strengths,
+ vulnerabilities, and likely outcomes.
+- **Recommendations** (if appropriate) for strategies that might help the individual achieve a better outcome.
+.  Base your analysis strictly on the information provided. If important information is missing or ambiguous, note the
+ limitations in your assessment.
+
+---
+# EXAMPLE 
+USER:
+***Psychodata***
+The subject is a 27 year old male.
+- He has poor impulse control and low level of patience. He lacks the ability to focus and/or commit to sustained challenges requiring effort.
+- He is ego driven to the point of narcissim, every criticism is a threat to his self esteem.
+- In his wors
+***challenge***
+While standing in line for the cashier in a grocery store, a rude customer cuts in line in front of the subject. 
--- a/data/patterns/recommend_yoga_practice/system.md
+++ b/data/patterns/recommend_yoga_practice/system.md
@@ -0,0 +1,40 @@
+# IDENTITY
+You are an experienced **yoga instructor and mindful living coach**. Your role is to guide users in a calm, clear, and compassionate manner. You will help them by following the stipulated steps:  
+
+# STEPS
+- Teach and provide practicing routines for **safe, effective yoga poses** (asana) with step-by-step guidance  
+- Help user build a **personalized sequences** suited to their experience level, goals, and any physical limitations  
+- Lead **guided meditations and relaxation exercises** that promote mindfulness and emotional balance  
+- Offer **holistic lifestyle advice** inspired by yogic principles—covering breathwork (pranayama), nutrition, sleep, posture, and daily wellbeing practices  
+- Foster an **atmosphere of serenity, self-awareness, and non-judgment** in every response  
+
+When responding, adapt your tone to be **soothing, encouraging, and introspective**, like a seasoned yoga teacher who integrates ancient wisdom into modern life.  
+
+# OUTPUT
+Use the following structure in your replies:  
+1. **Opening grounding statement** – a brief reflection or centering phrase.  
+2. **Main guidance** – offer detailed, safe, and clear instructions or insights relevant to the user’s query.  
+3. **Mindful takeaway** – close with a short reminder or reflection for continued mindfulness.  
+
+If users share specific goals (e.g., flexibility, relaxation, stress relief, back pain), **personalize** poses, sequences, or meditation practices accordingly.  
+
+If the user asks about a physical pose:  
+- Describe alignment carefully  
+- Explain how to modify for beginners or for safety  
+- Indicate common mistakes and how to avoid them  
+
+If the user asks about meditation or lifestyle:  
+- Offer simple, applicable techniques  
+- Encourage consistency and self-compassion  
+
+# EXAMPLE
+USER: Recommend a gentle yoga sequence for improving focus during stressful workdays.  
+
+Expected Output Example:  
+1. Begin with a short centering breath to quiet the mind.  
+2. Flow through seated side stretches, cat-cow, mountain pose, and standing forward fold. 
+3. Conclude with a brief meditation on the breath.  
+4. Reflect on how each inhale brings focus, and each exhale releases tension.  
+
+End every interaction with a phrase like:  
+> “Breathe in calm, breathe out ease.”
--- a/data/patterns/suggest_pattern/system.md
+++ b/data/patterns/suggest_pattern/system.md
@@ -73,21 +73,21 @@ Match the request to one or more of these primary categories:

 **AI**: ai, create_ai_jobs_analysis, create_art_prompt, create_pattern, create_prediction_block, extract_mcp_servers, extract_wisdom_agents, generate_code_rules, improve_prompt, judge_output, rate_ai_response, rate_ai_result, raw_query, suggest_pattern, summarize_prompt

-**ANALYSIS**: ai, analyze_answers, analyze_bill, analyze_bill_short, analyze_candidates, analyze_cfp_submission, analyze_claims, analyze_comments, analyze_debate, analyze_email_headers, analyze_incident, analyze_interviewer_techniques, analyze_logs, analyze_malware, analyze_military_strategy, analyze_mistakes, analyze_paper, analyze_paper_simple, analyze_patent, analyze_personality, analyze_presentation, analyze_product_feedback, analyze_proposition, analyze_prose, analyze_prose_json, analyze_prose_pinker, analyze_risk, analyze_sales_call, analyze_spiritual_text, analyze_tech_impact, analyze_terraform_plan, analyze_threat_report, analyze_threat_report_cmds, analyze_threat_report_trends, apply_ul_tags, check_agreement, compare_and_contrast, create_ai_jobs_analysis, create_idea_compass, create_investigation_visualization, create_prediction_block, create_recursive_outline, create_story_about_people_interaction, create_tags, dialog_with_socrates, extract_main_idea, extract_predictions, find_hidden_message, find_logical_fallacies, get_wow_per_minute, identify_dsrp_distinctions, identify_dsrp_perspectives, identify_dsrp_relationships, identify_dsrp_systems, identify_job_stories, label_and_rate, prepare_7s_strategy, provide_guidance, rate_content, rate_value, recommend_artists, recommend_talkpanel_topics, review_design, summarize_board_meeting, t_analyze_challenge_handling, t_check_dunning_kruger, t_check_metrics, t_describe_life_outlook, t_extract_intro_sentences, t_extract_panel_topics, t_find_blindspots, t_find_negative_thinking, t_red_team_thinking, t_threat_model_plans, t_year_in_review, write_hackerone_report
+**ANALYSIS**: ai, analyze_answers, analyze_bill, analyze_bill_short, analyze_candidates, analyze_cfp_submission, analyze_claims, analyze_comments, analyze_debate, analyze_email_headers, analyze_incident, analyze_interviewer_techniques, analyze_logs, analyze_malware, analyze_military_strategy, analyze_mistakes, analyze_paper, analyze_paper_simple, analyze_patent, analyze_personality, analyze_presentation, analyze_product_feedback, analyze_proposition, analyze_prose, analyze_prose_json, analyze_prose_pinker, analyze_risk, analyze_sales_call, analyze_spiritual_text, analyze_tech_impact, analyze_terraform_plan, analyze_threat_report, analyze_threat_report_cmds, analyze_threat_report_trends, apply_ul_tags, check_agreement, compare_and_contrast, concall_summary, create_ai_jobs_analysis, create_idea_compass, create_investigation_visualization, create_prediction_block, create_recursive_outline, create_story_about_people_interaction, create_tags, dialog_with_socrates, extract_main_idea, extract_predictions, find_hidden_message, find_logical_fallacies, get_wow_per_minute, identify_dsrp_distinctions, identify_dsrp_perspectives, identify_dsrp_relationships, identify_dsrp_systems, identify_job_stories, label_and_rate, model_as_sherlock_freud, predict_person_actions, prepare_7s_strategy, provide_guidance, rate_content, rate_value, recommend_artists, recommend_talkpanel_topics, review_design, summarize_board_meeting, t_analyze_challenge_handling, t_check_dunning_kruger, t_check_metrics, t_describe_life_outlook, t_extract_intro_sentences, t_extract_panel_topics, t_find_blindspots, t_find_negative_thinking, t_red_team_thinking, t_threat_model_plans, t_year_in_review, write_hackerone_report

 **BILL**: analyze_bill, analyze_bill_short

-**BUSINESS**: check_agreement, create_ai_jobs_analysis, create_formal_email, create_hormozi_offer, create_loe_document, create_logo, create_newsletter_entry, create_prd, explain_project, extract_business_ideas, extract_characters, extract_product_features, extract_skills, extract_sponsors, identify_job_stories, prepare_7s_strategy, rate_value, t_check_metrics, t_create_h3_career, t_visualize_mission_goals_projects, t_year_in_review, transcribe_minutes
+**BUSINESS**: check_agreement, concall_summary, create_ai_jobs_analysis, create_formal_email, create_hormozi_offer, create_loe_document, create_logo, create_newsletter_entry, create_prd, explain_project, extract_business_ideas, extract_characters, extract_product_features, extract_skills, extract_sponsors, identify_job_stories, prepare_7s_strategy, rate_value, t_check_metrics, t_create_h3_career, t_visualize_mission_goals_projects, t_year_in_review, transcribe_minutes

 **CLASSIFICATION**: apply_ul_tags

-**CONVERSION**: clean_text, convert_to_markdown, create_graph_from_input, export_data_as_csv, extract_videoid, get_youtube_rss, humanize, md_callout, sanitize_broken_html_to_markdown, to_flashcards, transcribe_minutes, translate, tweet, write_latex
+**CONVERSION**: clean_text, convert_to_markdown, create_graph_from_input, export_data_as_csv, extract_videoid, humanize, md_callout, sanitize_broken_html_to_markdown, to_flashcards, transcribe_minutes, translate, tweet, write_latex

 **CR THINKING**: capture_thinkers_work, create_idea_compass, create_markmap_visualization, dialog_with_socrates, extract_alpha, extract_controversial_ideas, extract_extraordinary_claims, extract_predictions, extract_primary_problem, extract_wisdom_nometa, find_hidden_message, find_logical_fallacies, summarize_debate, t_analyze_challenge_handling, t_check_dunning_kruger, t_find_blindspots, t_find_negative_thinking, t_find_neglected_goals, t_red_team_thinking

 **CREATIVITY**: create_mnemonic_phrases, write_essay

-**DEVELOPMENT**: agility_story, analyze_logs, analyze_prose_json, answer_interview_question, ask_secure_by_design_questions, ask_uncle_duke, coding_master, create_coding_feature, create_coding_project, create_command, create_design_document, create_git_diff_commit, create_loe_document, create_mermaid_visualization, create_mermaid_visualization_for_github, create_pattern, create_prd, create_sigma_rules, create_user_story, explain_code, explain_docs, explain_project, export_data_as_csv, extract_algorithm_update_recommendations, extract_mcp_servers, extract_poc, extract_product_features, generate_code_rules, get_youtube_rss, identify_job_stories, improve_prompt, official_pattern_template, recommend_pipeline_upgrades, refine_design_document, review_code, review_design, sanitize_broken_html_to_markdown, suggest_pattern, summarize_git_changes, summarize_git_diff, summarize_pull-requests, write_nuclei_template_rule, write_pull-request, write_semgrep_rule
+**DEVELOPMENT**: agility_story, analyze_logs, analyze_prose_json, answer_interview_question, ask_secure_by_design_questions, ask_uncle_duke, coding_master, create_coding_feature, create_coding_project, create_command, create_design_document, create_git_diff_commit, create_loe_document, create_mermaid_visualization, create_mermaid_visualization_for_github, create_pattern, create_prd, create_sigma_rules, create_user_story, explain_code, explain_docs, explain_project, export_data_as_csv, extract_algorithm_update_recommendations, extract_mcp_servers, extract_poc, extract_product_features, generate_code_rules, identify_job_stories, improve_prompt, official_pattern_template, recommend_pipeline_upgrades, refine_design_document, review_code, review_design, sanitize_broken_html_to_markdown, suggest_pattern, summarize_git_changes, summarize_git_diff, summarize_pull-requests, write_nuclei_template_rule, write_pull-request, write_semgrep_rule

 **DEVOPS**: analyze_terraform_plan

@@ -105,17 +105,19 @@ Match the request to one or more of these primary categories:

 **SECURITY**: analyze_email_headers, analyze_incident, analyze_logs, analyze_malware, analyze_risk, analyze_terraform_plan, analyze_threat_report, analyze_threat_report_cmds, analyze_threat_report_trends, ask_secure_by_design_questions, create_command, create_cyber_summary, create_graph_from_input, create_investigation_visualization, create_network_threat_landscape, create_report_finding, create_security_update, create_sigma_rules, create_stride_threat_model, create_threat_scenarios, create_ttrc_graph, create_ttrc_narrative, extract_ctf_writeup, improve_report_finding, recommend_pipeline_upgrades, review_code, t_red_team_thinking, t_threat_model_plans, write_hackerone_report, write_nuclei_template_rule, write_semgrep_rule

-**SELF**: analyze_mistakes, analyze_personality, analyze_spiritual_text, create_better_frame, create_diy, create_reading_plan, create_story_about_person, dialog_with_socrates, extract_article_wisdom, extract_book_ideas, extract_book_recommendations, extract_insights, extract_insights_dm, extract_most_redeeming_thing, extract_recipe, extract_recommendations, extract_song_meaning, extract_wisdom, extract_wisdom_dm, extract_wisdom_short, find_female_life_partner, heal_person, provide_guidance, recommend_artists, t_check_dunning_kruger, t_create_h3_career, t_describe_life_outlook, t_find_neglected_goals, t_give_encouragement
+**SELF**: analyze_mistakes, analyze_personality, analyze_spiritual_text, create_better_frame, create_diy, create_reading_plan, create_story_about_person, dialog_with_socrates, extract_article_wisdom, extract_book_ideas, extract_book_recommendations, extract_insights, extract_insights_dm, extract_most_redeeming_thing, extract_recipe, extract_recommendations, extract_song_meaning, extract_wisdom, extract_wisdom_dm, extract_wisdom_short, find_female_life_partner, heal_person, model_as_sherlock_freud, predict_person_actions, provide_guidance, recommend_artists, recommend_yoga_practice, t_check_dunning_kruger, t_create_h3_career, t_describe_life_outlook, t_find_neglected_goals, t_give_encouragement

 **STRATEGY**: analyze_military_strategy, create_better_frame, prepare_7s_strategy, t_analyze_challenge_handling, t_find_blindspots, t_find_negative_thinking, t_find_neglected_goals, t_red_team_thinking, t_threat_model_plans, t_visualize_mission_goals_projects

-**SUMMARIZE**: capture_thinkers_work, create_5_sentence_summary, create_micro_summary, create_newsletter_entry, create_show_intro, create_summary, extract_core_message, extract_latest_video, extract_main_idea, summarize, summarize_board_meeting, summarize_debate, summarize_git_changes, summarize_git_diff, summarize_lecture, summarize_legislation, summarize_meeting, summarize_micro, summarize_newsletter, summarize_paper, summarize_pull-requests, summarize_rpg_session, youtube_summary
+**SUMMARIZE**: capture_thinkers_work, concall_summary, create_5_sentence_summary, create_micro_summary, create_newsletter_entry, create_show_intro, create_summary, extract_core_message, extract_latest_video, extract_main_idea, summarize, summarize_board_meeting, summarize_debate, summarize_git_changes, summarize_git_diff, summarize_lecture, summarize_legislation, summarize_meeting, summarize_micro, summarize_newsletter, summarize_paper, summarize_pull-requests, summarize_rpg_session, youtube_summary

-**VISUALIZE**: create_excalidraw_visualization, create_graph_from_input, create_idea_compass, create_investigation_visualization, create_keynote, create_logo, create_markmap_visualization, create_mermaid_visualization, create_mermaid_visualization_for_github, create_video_chapters, create_visualization, enrich_blog_post, t_visualize_mission_goals_projects
+**VISUALIZE**: create_conceptmap, create_excalidraw_visualization, create_graph_from_input, create_idea_compass, create_investigation_visualization, create_keynote, create_logo, create_markmap_visualization, create_mermaid_visualization, create_mermaid_visualization_for_github, create_video_chapters, create_visualization, enrich_blog_post, t_visualize_mission_goals_projects

 **WISDOM**: extract_alpha, extract_article_wisdom, extract_book_ideas, extract_insights, extract_most_redeeming_thing, extract_recommendations, extract_wisdom, extract_wisdom_dm, extract_wisdom_nometa, extract_wisdom_short

-**WRITING**: analyze_prose_json, analyze_prose_pinker, apply_ul_tags, clean_text, compare_and_contrast, convert_to_markdown, create_5_sentence_summary, create_academic_paper, create_aphorisms, create_better_frame, create_design_document, create_diy, create_formal_email, create_hormozi_offer, create_keynote, create_micro_summary, create_newsletter_entry, create_prediction_block, create_prd, create_show_intro, create_story_about_people_interaction, create_story_explanation, create_summary, create_tags, create_user_story, enrich_blog_post, explain_docs, explain_terms, humanize, improve_academic_writing, improve_writing, label_and_rate, md_callout, official_pattern_template, recommend_talkpanel_topics, refine_design_document, summarize, summarize_debate, summarize_lecture, summarize_legislation, summarize_meeting, summarize_micro, summarize_newsletter, summarize_paper, summarize_rpg_session, t_create_opening_sentences, t_describe_life_outlook, t_extract_intro_sentences, t_extract_panel_topics, t_give_encouragement, t_year_in_review, transcribe_minutes, tweet, write_essay, write_essay_pg, write_hackerone_report, write_latex, write_micro_essay, write_pull-request
+**WELLNESS**: analyze_spiritual_text, create_better_frame, extract_wisdom_dm, heal_person, model_as_sherlock_freud, predict_person_actions, provide_guidance, recommend_yoga_practice, t_give_encouragement
+
+**WRITING**: analyze_prose_json, analyze_prose_pinker, apply_ul_tags, clean_text, compare_and_contrast, convert_to_markdown, create_5_sentence_summary, create_academic_paper, create_aphorisms, create_better_frame, create_design_document, create_diy, create_formal_email, create_hormozi_offer, create_keynote, create_micro_summary, create_newsletter_entry, create_prediction_block, create_prd, create_show_intro, create_story_about_people_interaction, create_story_explanation, create_summary, create_tags, create_user_story, enrich_blog_post, explain_docs, explain_terms, fix_typos, humanize, improve_academic_writing, improve_writing, label_and_rate, md_callout, official_pattern_template, recommend_talkpanel_topics, refine_design_document, summarize, summarize_debate, summarize_lecture, summarize_legislation, summarize_meeting, summarize_micro, summarize_newsletter, summarize_paper, summarize_rpg_session, t_create_opening_sentences, t_describe_life_outlook, t_extract_intro_sentences, t_extract_panel_topics, t_give_encouragement, t_year_in_review, transcribe_minutes, tweet, write_essay, write_essay_pg, write_hackerone_report, write_latex, write_micro_essay, write_pull-request

 ## Workflow Suggestions

--- a/data/patterns/suggest_pattern/user.md
+++ b/data/patterns/suggest_pattern/user.md
@@ -196,6 +196,10 @@ Review contract to identify stipulations, issues, and changes for negotiation.

 Create comparisons table, highlighting key differences and similarities.

+### concall_summary
+
+Analyze earnings call transcripts to extract management insights, financial metrics, and investment implications.
+
 ### create_ai_jobs_analysis

 Identify automation risks and career resilience strategies.
@@ -296,6 +300,14 @@ Extract/analyze user job stories to understand motivations.

 Categorize/evaluate content by assigning labels and ratings.

+### model_as_sherlock_freud
+
+Builds psychological models using detective reasoning and psychoanalytic insight.
+
+### predict_person_actions
+
+Predicts behavioral responses based on psychological profiles and challenges
+
 ### prepare_7s_strategy

 Apply McKinsey 7S framework to analyze organizational alignment.
@@ -598,6 +610,10 @@ Transform technical docs into clearer explanations with examples.

 Create glossaries of advanced terms with definitions and analogies.

+### fix_typos
+
+Proofreads and corrects typos, spelling, grammar, and punctuation errors.
+
 ### humanize

 Transform technical content into approachable language.
@@ -880,6 +896,10 @@ Convert content into flashcard format for learning.

 ## VISUALIZATION PATTERNS

+### create_conceptmap
+
+Transform unstructured text or markdown content into interactive HTML concept maps using Vis.js by extracting key concepts and their logical relationships.
+
 ### create_excalidraw_visualization

 Create visualizations using Excalidraw.
@@ -926,10 +946,6 @@ Convert content to markdown, preserving original content and structure.

 Extract data and convert to CSV, preserving data integrity.

-### get_youtube_rss
-
-Generate RSS feed URLs for YouTube channels.
-
 ### sanitize_broken_html_to_markdown

 Clean/convert malformed HTML to markdown.
@@ -983,3 +999,9 @@ Summarize RPG sessions capturing events, combat, and narrative.
 ### extract_jokes

 Extract/categorize jokes, puns, and witty remarks.
+
+## WELLNESS PATTERNS
+
+### recommend_yoga_practice
+
+Provides personalized yoga sequences, meditation guidance, and holistic lifestyle advice based on individual profiles.
--- a/docs/CONTRIBUTING.md
+++ b/docs/CONTRIBUTING.md
@@ -51,6 +51,29 @@ docs: update installation instructions

 ## Pull Request Process

+### Pull Request Guidelines
+
+**Keep pull requests focused and minimal.**
+
+PRs that touch a large number of files (50+) without clear functional justification will likely be rejected without detailed review.
+
+#### Why we enforce this
+
+- **Reviewability**: Large PRs are effectively un-reviewable. Studies show reviewer effectiveness drops significantly after ~200-400 lines of code. A 93-file "cleanup" PR cannot receive meaningful review.
+- **Git history**: Sweeping changes pollute `git blame`, making it harder to trace when and why functional changes were made.
+- **Merge conflicts**: Large PRs increase the likelihood of conflicts with other contributors' work.
+- **Risk**: More changed lines means more opportunities for subtle bugs, even in "safe" refactors.
+
+#### What to do instead
+
+If you have a large change in mind, break it into logical, independently-mergeable slices. For example:
+
+- ✅ "Replace `interface{}` with `any` across codebase" (single mechanical change, easy to verify)
+- ✅ "Migrate to `strings.CutPrefix` in `internal/cli`" (scoped to one package)
+- ❌ "Modernize codebase with multiple idiom updates" (too broad, impossible to review)
+
+For sweeping refactors or style changes, **open an issue first** to discuss the approach with maintainers before investing time in the work.
+
 ### Changelog Generation (REQUIRED)

 After opening your PR, generate a changelog entry:
--- a/docs/GitHub-Models-Setup.md
+++ b/docs/GitHub-Models-Setup.md
@@ -0,0 +1,700 @@
+# GitHub Models Setup Guide for Fabric
+
+This guide will walk you through setting up and using GitHub Models with Fabric CLI. GitHub Models provides free access to multiple AI models from OpenAI, Meta, Microsoft, DeepSeek, xAI, and other providers using only your GitHub credentials.
+
+## Table of Contents
+
+- [What are GitHub Models?](#what-are-github-models)
+- [Getting Your GitHub Models API Key](#getting-your-github-models-api-key)
+- [Configuring Fabric for GitHub Models](#configuring-fabric-for-github-models)
+- [Testing Your Setup](#testing-your-setup)
+- [Available Models](#available-models)
+- [Rate Limits & Free Tier](#rate-limits--free-tier)
+- [Troubleshooting](#troubleshooting)
+- [Advanced Usage](#advanced-usage)
+
+---
+
+## What are GitHub Models?
+
+**GitHub Models** is a free AI inference API platform that allows you to access multiple AI models using only your GitHub account. It's powered by Azure AI infrastructure and provides:
+
+- **Unified Access**: Single API endpoint for models from multiple providers
+- **No Extra API Keys**: Uses GitHub Personal Access Tokens (no separate OpenAI, Anthropic, etc. keys needed)
+- **Free Tier**: Rate-limited free access perfect for prototyping and personal projects
+- **Web Playground**: Test models directly at [github.com/marketplace/models](https://github.com/marketplace/models)
+- **Compatible Format**: Works with OpenAI SDK standards
+
+### Why Use GitHub Models with Fabric?
+
+- **No Cost for Testing**: Free tier allows 50-150 requests/day depending on model
+- **Multiple Providers**: Access OpenAI, Meta Llama, Microsoft Phi, DeepSeek, and more
+- **Easy Setup**: Just one GitHub token instead of managing multiple API keys
+- **Great for Learning**: Experiment with different models without financial commitment
+
+---
+
+## Getting Your GitHub Models API Key
+
+GitHub Models uses **Personal Access Tokens (PAT)** instead of separate API keys.
+
+### Step-by-Step Instructions
+
+1. **Sign in to GitHub** at [github.com](https://github.com)
+
+2. **Navigate to Token Settings:**
+   - Click your profile picture (upper-right corner)
+   - Click **Settings**
+   - Scroll down the left sidebar to **Developer settings** (at the bottom)
+   - Click **Personal access tokens** → **Fine-grained tokens** (recommended)
+
+3. **Generate New Token:**
+   - Click **Generate new token**
+   - Give it a descriptive name: `Fabric CLI - GitHub Models`
+   - Set expiration (recommended: 90 days or custom)
+   - **Repository access**: Select "Public Repositories (read-only)" or "All repositories" (your choice)
+   - **Permissions**:
+     - Scroll down to **Account permissions**
+     - Find **AI Models** and set to **Read-only** ✓
+     - This grants the `models:read` scope
+   - Click **Generate token** at the bottom
+
+4. **Save Your Token:**
+   - **IMPORTANT**: Copy the token immediately (starts with `github_pat_` or `ghp_`)
+   - You won't be able to see it again!
+   - Store it securely - this will be your `GITHUB_TOKEN`
+
+### Security Best Practices
+
+- ✅ Use fine-grained tokens with minimal permissions
+- ✅ Set an expiration date (rotate tokens regularly)
+- ✅ Never commit tokens to Git repositories
+- ✅ Store in environment variables or secure credential managers
+- ❌ Don't share tokens in chat, email, or screenshots
+
+---
+
+## Configuring Fabric for GitHub Models
+
+### Method 1: Using Fabric Setup (Recommended)
+
+This is the easiest and safest method:
+
+1. **Run Fabric Setup:**
+
+   ```bash
+   fabric --setup
+   ```
+
+2. **Select GitHub from the Menu:**
+   - You'll see a numbered list of AI vendors
+   - Find `[8] GitHub (configured)` or similar
+   - Enter the number (e.g., `8`) and press Enter
+
+3. **Enter Your GitHub Token:**
+   - When prompted for "API Key", paste your GitHub Personal Access Token
+   - The token you created earlier (starts with `github_pat_` or `ghp_`)
+   - Press Enter
+
+4. **Verify Base URL (Optional):**
+   - You'll be asked for "API Base URL"
+   - Press Enter to use the default: `https://models.github.ai/inference`
+   - Or customize if needed (advanced use only)
+
+5. **Save and Exit:**
+   - The setup wizard will save your configuration
+   - You should see "GitHub (configured)" next time
+
+### Method 2: Manual Configuration (Advanced)
+
+If you prefer to manually edit the configuration file:
+
+1. **Edit Environment File:**
+
+   ```bash
+   nano ~/.config/fabric/.env
+   ```
+
+2. **Add GitHub Configuration:**
+
+   ```bash
+   # GitHub Models API Key (your Personal Access Token)
+   GITHUB_API_KEY=github_pat_YOUR_TOKEN_HERE
+
+   # GitHub Models API Base URL (default, usually don't need to change)
+   GITHUB_API_BASE_URL=https://models.github.ai/inference
+   ```
+
+   Save and exit (Ctrl+X, then Y, then Enter)
+
+**Note**: The environment variable is `GITHUB_API_KEY`, not `GITHUB_TOKEN`.
+
+### Verify Configuration
+
+Check that your configuration is properly set:
+
+```bash
+grep GITHUB_API_KEY ~/.config/fabric/.env
+```
+
+You should see:
+
+```text
+GITHUB_API_KEY=github_pat_...
+```
+
+Or run setup again to verify:
+
+```bash
+fabric --setup
+```
+
+Look for `[8] GitHub (configured)` in the list.
+
+---
+
+## Testing Your Setup
+
+### 1. List Available Models
+
+Verify that Fabric can connect to GitHub Models and fetch the model list:
+
+```bash
+fabric --listmodels | grep GitHub
+```
+
+**Expected Output:**
+
+```text
+Available models:
+...
+$ fabric -L | grep GitHub
+        [65]    GitHub|ai21-labs/ai21-jamba-1.5-large
+        [66]    GitHub|cohere/cohere-command-a
+        [67]    GitHub|cohere/cohere-command-r-08-2024
+        [68]    GitHub|cohere/cohere-command-r-plus-08-2024
+        [69]    GitHub|deepseek/deepseek-r1
+        [70]    GitHub|deepseek/deepseek-r1-0528
+        [71]    GitHub|deepseek/deepseek-v3-0324
+        [72]    GitHub|meta/llama-3.2-11b-vision-instruct
+        [73]    GitHub|meta/llama-3.2-90b-vision-instruct
+... (and more)
+```
+
+### 2. Simple Chat Test
+
+Test a basic chat completion with a small, fast model:
+
+```bash
+# Use gpt-4o-mini (fast and has generous rate limits)
+fabric --vendor GitHub -m openai/gpt-4o-mini 'Why is th
+e sky blue?'
+```
+
+**Expected**: You should see a response explaining Rayleigh scattering.
+
+**Tip**: Model names from `--listmodels` can be used directly (e.g., `openai/gpt-4o-mini`, `openai/gpt-4o`, `meta/llama-4-maverick-17b-128e-instruct-fp8`).
+
+### 3. Test with a Pattern
+
+Use one of Fabric's built-in patterns:
+
+```bash
+echo "Artificial intelligence is transforming how we work and live." | \
+  fabric --pattern summarize --vendor GitHub --model "openai/gpt-4o-mini"
+```
+
+### 4. Test Streaming
+
+Verify streaming responses work:
+
+```bash
+echo "Count from 1 to 100" | \
+  fabric --vendor GitHub --model "openai/gpt-4o-mini" --stream
+```
+
+You should see the response appear progressively, word by word.
+
+### 5. Test with Different Models
+
+Try a Meta Llama model:
+
+```bash
+# Use a Llama model
+echo "Explain quantum computing" | \
+  fabric --vendor GitHub --model "meta/Meta-Llama-3.1-8B-Instruct"
+```
+
+### Quick Validation Checklist
+
+- [x] `--listmodels` shows GitHub models
+- [x] Basic chat completion works
+- [x] Patterns work with GitHub vendor
+- [x] Streaming responses work
+- [x] Can switch between different models
+
+---
+
+## Available Models
+
+GitHub Models provides access to models from multiple providers. Models use the format: `{publisher}/{model-name}`
+
+### OpenAI Models
+
+| Model ID | Description | Tier | Best For |
+|----------|-------------|------|----------|
+| `openai/gpt-4.1` | Latest flagship GPT-4 | High | Complex tasks, reasoning |
+| `openai/gpt-4o` | Optimized GPT-4 | High | General purpose, fast |
+| `openai/gpt-4o-mini` | Compact, cost-effective | Low | Quick tasks, high volume |
+| `openai/o1` | Advanced reasoning | High | Complex problem solving |
+| `openai/o3` | Next-gen reasoning | High | Cutting-edge reasoning |
+
+### Meta Llama Models
+
+| Model ID | Description | Tier | Best For |
+|----------|-------------|------|----------|
+| `meta/llama-3.1-405b` | Largest Llama model | High | Complex tasks, accuracy |
+| `meta/llama-3.1-70b` | Mid-size Llama | Low | Balanced performance |
+| `meta/llama-3.1-8b` | Compact Llama | Low | Fast, efficient tasks |
+
+### Microsoft Phi Models
+
+| Model ID | Description | Tier | Best For |
+|----------|-------------|------|----------|
+| `microsoft/phi-4` | Latest Phi generation | Low | Efficient reasoning |
+| `microsoft/phi-3-medium` | Mid-size variant | Low | General tasks |
+| `microsoft/phi-3-mini` | Smallest Phi | Low | Quick, simple tasks |
+
+### DeepSeek Models
+
+| Model ID | Description | Tier | Special |
+|----------|-------------|------|---------|
+| `deepseek/deepseek-r1` | Reasoning model | Very Limited | 8 requests/day |
+| `deepseek/deepseek-r1-0528` | Updated version | Very Limited | 8 requests/day |
+
+### xAI Models
+
+| Model ID | Description | Tier | Special |
+|----------|-------------|------|---------|
+| `xai/grok-3` | Latest Grok | Very Limited | 15 requests/day |
+| `xai/grok-3-mini` | Smaller Grok | Very Limited | 15 requests/day |
+
+### Getting the Full List
+
+To see all currently available models:
+
+```bash
+fabric --listmodels | grep GitHub
+```
+
+Or for a formatted list with details, you can query the GitHub Models API directly:
+
+```bash
+curl -H "Authorization: Bearer $GITHUB_TOKEN" \
+     -H "X-GitHub-Api-Version: 2022-11-28" \
+     https://models.github.ai/catalog/models | jq '.[] | {id, publisher, tier: .rate_limit_tier}'
+```
+
+---
+
+## Rate Limits & Free Tier
+
+GitHub Models has tiered rate limits based on model complexity. Understanding these helps you use the free tier effectively.
+
+### Low Tier Models (Recommended for High Volume)
+
+**Models**: `gpt-4o-mini`, `llama-3.1-*`, `phi-*`
+
+- **Requests per minute**: 15
+- **Requests per day**: 150
+- **Tokens per request**: 8,000 input / 4,000 output
+- **Concurrent requests**: 5
+
+**Best practices**: Use these for most Fabric patterns and daily tasks.
+
+### High Tier Models (Use Sparingly)
+
+**Models**: `gpt-4.1`, `gpt-4o`, `o1`, `o3`, `llama-3.1-405b`
+
+- **Requests per minute**: 10
+- **Requests per day**: 50
+- **Tokens per request**: 8,000 input / 4,000 output
+- **Concurrent requests**: 2
+
+**Best practices**: Save for complex tasks, important queries, or when you need maximum quality.
+
+### Very Limited Models
+
+**Models**: `deepseek-r1`, `grok-3`
+
+- **Requests per minute**: 1
+- **Requests per day**: 8-15 (varies by model)
+- **Tokens per request**: 4,000 input / 4,000 output
+- **Concurrent requests**: 1
+
+**Best practices**: Use only for special experiments or when you specifically need these models.
+
+### Rate Limit Reset Times
+
+- **Per-minute limits**: Reset every 60 seconds
+- **Daily limits**: Reset at midnight UTC
+- **Per-user**: Limits are tied to your GitHub account, not the token
+
+### Enhanced Limits with GitHub Copilot
+
+If you have a GitHub Copilot subscription, you get higher limits:
+
+- **Copilot Business**: 2× daily request limits
+- **Copilot Enterprise**: 3× daily limits + higher token limits
+
+### What Happens When You Hit Limits?
+
+You'll receive an HTTP 429 error with a message like:
+
+```text
+Rate limit exceeded. Try again in X seconds.
+```
+
+Fabric will display this error. Wait for the reset time and try again.
+
+### Tips for Staying Within Limits
+
+1. **Use low-tier models** for most tasks (`gpt-4o-mini`, `llama-3.1-8b`)
+2. **Batch your requests** - process multiple items together when possible
+3. **Cache results** - save responses for repeated queries
+4. **Monitor usage** - keep track of daily request counts
+5. **Set per-pattern models** - configure specific models for specific patterns (see Advanced Usage)
+
+---
+
+## Troubleshooting
+
+### Error: "Authentication failed" or "Unauthorized"
+
+**Cause**: Invalid or missing GitHub token
+
+**Solutions**:
+
+1. Verify token is in `.env` file:
+
+   ```bash
+   grep GITHUB_API_KEY ~/.config/fabric/.env
+   ```
+
+2. Check token has `models:read` permission:
+   - Go to GitHub Settings → Developer settings → Personal access tokens
+   - Click on your token
+   - Verify "AI Models: Read-only" is checked
+
+3. Re-run setup to reconfigure:
+
+   ```bash
+   fabric --setup
+   # Select GitHub (number 8 or similar)
+   # Enter your token again
+   ```
+
+4. Generate a new token if needed (tokens expire)
+
+### Error: "Rate limit exceeded"
+
+**Cause**: Too many requests in a short time period
+
+**Solutions**:
+
+1. Check which tier your model is in (see [Rate Limits](#rate-limits--free-tier))
+2. Wait for the reset (check error message for wait time)
+3. Switch to a lower-tier model:
+
+   ```bash
+   # Instead of gpt-4.1 (high tier)
+   fabric --vendor GitHub --model openai/gpt-4.1 ...
+
+   # Use gpt-4o-mini (low tier)
+   fabric --vendor GitHub --model openai/gpt-4o-mini ...
+   ```
+
+### Error: "Model not found" or "Invalid model"
+
+**Cause**: Model name format incorrect or model not available
+
+**Solutions**:
+
+1. Use correct format: `{publisher}/{model-name}`, e.g., `openai/gpt-4o-mini`
+
+   ```bash
+   # ❌ Wrong
+   fabric --vendor GitHub --model gpt-4o-mini
+
+   # ✅ Correct
+   fabric --vendor GitHub --model openai/gpt-4o-mini
+   ```
+
+2. List available models to verify name:
+
+   ```bash
+   fabric --listmodels --vendor GitHub | grep -i "gpt-4"
+   ```
+
+### Error: "Cannot list models" or Empty model list
+
+**Cause**: API endpoint issue or authentication problem
+
+**Solutions**:
+
+1. Test direct API access:
+
+   ```bash
+   curl -H "Authorization: Bearer $GITHUB_TOKEN" \
+        -H "X-GitHub-Api-Version: 2022-11-28" \
+        https://models.github.ai/catalog/models
+   ```
+
+2. If curl works but Fabric doesn't, rebuild Fabric:
+
+   ```bash
+   cd /path/to/fabric
+   go build ./cmd/fabric
+   ```
+
+3. Check for network/firewall issues blocking `models.github.ai`
+
+### Error: "Response format not supported"
+
+**Cause**: This should be fixed in the latest version with direct fetch fallback
+
+**Solutions**:
+
+1. Update to the latest Fabric version with PR #1839 merged
+2. Verify you're on a version that includes the `FetchModelsDirectly` fallback
+
+### Models are slow to respond
+
+**Cause**: High tier models have limited concurrency, or GitHub Models API congestion
+
+**Solutions**:
+
+1. Switch to faster models:
+   - `openai/gpt-4o-mini` instead of `gpt-4.1`
+   - `meta/llama-3.1-8b` instead of `llama-3.1-405b`
+
+2. Check your internet connection
+
+3. Try again later (API may be experiencing high traffic)
+
+### Token expires or becomes invalid
+
+**Cause**: Tokens have expiration dates or can be revoked
+
+**Solutions**:
+
+1. Generate a new token (see [Getting Your GitHub Models API Key](#getting-your-github-models-api-key))
+2. Update `.env` file with new token
+3. Set longer expiration when creating tokens (e.g., 90 days)
+
+---
+
+## Advanced Usage
+
+### Using Specific Models with Patterns
+
+You can specify which model to use with any pattern:
+
+```bash
+# Use GPT-4.1 with the analyze_claims pattern
+cat article.txt | fabric --pattern analyze_claims \
+  --vendor GitHub --model openai/gpt-4.1
+
+# Use Llama for summarization
+cat document.txt | fabric --pattern summarize \
+  --vendor GitHub --model meta/llama-3.1-70b
+```
+
+### Per-Pattern Model Mapping
+
+Set default models for specific patterns using environment variables:
+
+Edit `~/.config/fabric/.env`:
+
+```bash
+# Use GPT-4.1 for complex analysis
+FABRIC_MODEL_analyze_claims=GitHub|openai/gpt-4.1
+FABRIC_MODEL_extract_wisdom=GitHub|openai/gpt-4.1
+
+# Use GPT-4o-mini for simple tasks
+FABRIC_MODEL_summarize=GitHub|openai/gpt-4o-mini
+FABRIC_MODEL_extract_article_wisdom=GitHub|openai/gpt-4o-mini
+
+# Use Llama for code tasks
+FABRIC_MODEL_explain_code=GitHub|meta/llama-3.1-70b
+```
+
+Now when you run:
+
+```bash
+cat article.txt | fabric --pattern analyze_claims
+```
+
+It will automatically use `GitHub|openai/gpt-4.1` without needing to specify the vendor and model.
+
+### Comparing Responses Across Providers
+
+Compare how different models respond to the same input:
+
+```bash
+# OpenAI GPT-4o-mini
+echo "Explain quantum computing" | \
+  fabric --vendor GitHub --model openai/gpt-4o-mini > response_openai.txt
+
+# Meta Llama
+echo "Explain quantum computing" | \
+  fabric --vendor GitHub --model meta/llama-3.1-70b > response_llama.txt
+
+# Microsoft Phi
+echo "Explain quantum computing" | \
+  fabric --vendor GitHub --model microsoft/phi-4 > response_phi.txt
+
+# Compare
+diff response_openai.txt response_llama.txt
+```
+
+### Testing Different Models for a Pattern
+
+Find the best model for your use case:
+
+```bash
+# Create a test script
+cat > test_models.sh << 'EOF'
+#!/bin/bash
+
+INPUT="Explain the concept of recursion in programming"
+PATTERN="explain_code"
+
+for MODEL in "openai/gpt-4o-mini" "meta/llama-3.1-8b" "microsoft/phi-4"; do
+  echo "=== Testing $MODEL ==="
+  echo "$INPUT" | fabric --pattern "$PATTERN" --vendor GitHub --model "$MODEL"
+  echo ""
+done
+EOF
+
+chmod +x test_models.sh
+./test_models.sh
+```
+
+### Quick Test Without Setup
+
+If you want to quickly test without running full setup, you can set the environment variable directly:
+
+```bash
+# Temporary test (this session only)
+export GITHUB_API_KEY=github_pat_YOUR_TOKEN_HERE
+
+# Test immediately
+fabric --listmodels --vendor GitHub
+```
+
+This is useful for quick tests, but we recommend using `fabric --setup` for permanent configuration.
+
+### Streaming for Long Responses
+
+For long-form content, use streaming to see results as they generate:
+
+```bash
+cat long_article.txt | \
+  fabric --pattern summarize \
+  --vendor GitHub --model openai/gpt-4o-mini \
+  --stream
+```
+
+### Saving Token Usage
+
+Monitor your usage to stay within rate limits:
+
+```bash
+# Create a simple usage tracker
+echo "$(date): Used gpt-4.1 for analyze_claims" >> ~/.config/fabric/usage.log
+
+# Check daily usage
+grep "$(date +%Y-%m-%d)" ~/.config/fabric/usage.log | wc -l
+```
+
+### Environment-Based Configuration
+
+Create different profiles for different use cases:
+
+```bash
+# Development profile (uses free GitHub Models)
+cat > ~/.config/fabric/.env.dev << EOF
+GITHUB_TOKEN=github_pat_dev_token_here
+DEFAULT_VENDOR=GitHub
+DEFAULT_MODEL=openai/gpt-4o-mini
+EOF
+
+# Production profile (uses paid OpenAI)
+cat > ~/.config/fabric/.env.prod << EOF
+OPENAI_API_KEY=sk-prod-key-here
+DEFAULT_VENDOR=OpenAI
+DEFAULT_MODEL=gpt-4
+EOF
+
+# Switch profiles
+ln -sf ~/.config/fabric/.env.dev ~/.config/fabric/.env
+```
+
+---
+
+## Additional Resources
+
+### Official Documentation
+
+- [GitHub Models Quickstart](https://docs.github.com/en/github-models/quickstart)
+- [GitHub Models API Reference](https://docs.github.com/en/rest/models)
+- [GitHub Models Marketplace](https://github.com/marketplace/models)
+
+### Fabric Documentation
+
+- [Fabric README](../README.md)
+- [Contexts and Sessions Tutorial](./contexts-and-sessions-tutorial.md)
+- [Using Speech-to-Text](./Using-Speech-To-Text.md)
+
+### Community
+
+- [Fabric GitHub Repository](https://github.com/danielmiessler/fabric)
+- [Fabric Issues](https://github.com/danielmiessler/fabric/issues)
+- [Fabric Discussions](https://github.com/danielmiessler/fabric/discussions)
+
+---
+
+## Summary
+
+GitHub Models provides an excellent way to experiment with AI models through Fabric without managing multiple API keys or incurring costs. Key points:
+
+✅ **Free to start**: No credit card required, 50-150 requests/day
+✅ **Multiple providers**: OpenAI, Meta, Microsoft, DeepSeek, xAI
+✅ **Simple setup**: Just one GitHub token via `fabric --setup`
+✅ **Great for learning**: Try different models and patterns
+✅ **Production path**: Can upgrade to paid tier when ready
+
+### Quick Start Commands
+
+```bash
+# 1. Get GitHub token with models:read scope from:
+#    https://github.com/settings/tokens
+
+# 2. Configure Fabric
+fabric --setup
+# Select [8] GitHub
+# Paste your token when prompted
+
+# 3. List available models
+fabric --listmodels --vendor GitHub | grep gpt-4o
+
+# 4. Try it out with gpt-4o-mini
+echo "What is AI?" | fabric --vendor GitHub --model "gpt-4o-mini"
+```
+
+**Recommended starting point**: Use `gpt-4o-mini` for most patterns - it's fast, capable, and has generous rate limits (150 requests/day).
+
+**Available Models**: `gpt-4o`, `gpt-4o-mini`, `Meta-Llama-3.1-8B-Instruct`, `Meta-Llama-3.1-70B-Instruct`, `Mistral-large-2407`, and more. Use `--listmodels` to see the complete list.
+
+Happy prompting! 🚀
--- a/flake.lock
+++ b/flake.lock
@@ -5,11 +5,11 @@
        "systems": "systems"
      },
      "locked": {
-        "lastModified": 1694529238,
-        "narHash": "sha256-zsNZZGTGnMOf9YpHKJqMSsa0dXbfmxeoJ7xHlrt+xmY=",
+        "lastModified": 1731533236,
+        "narHash": "sha256-l0KFg5HjrsfsO/JpG+r7fRrqm12kzFHyUHqHCVpMMbI=",
        "owner": "numtide",
        "repo": "flake-utils",
-        "rev": "ff7b65b44d01cf9ba6a71320833626af21126384",
+        "rev": "11707dc2f618dd54ca8739b309ec4fc024de578b",
        "type": "github"
      },
      "original": {
@@ -26,11 +26,11 @@
        ]
      },
      "locked": {
-        "lastModified": 1742209644,
-        "narHash": "sha256-jMy1XqXqD0/tJprEbUmKilTkvbDY/C0ZGSsJJH4TNCE=",
+        "lastModified": 1763982521,
+        "narHash": "sha256-ur4QIAHwgFc0vXiaxn5No/FuZicxBr2p0gmT54xZkUQ=",
        "owner": "nix-community",
        "repo": "gomod2nix",
-        "rev": "8f3534eb8f6c5c3fce799376dc3b91bae6b11884",
+        "rev": "02e63a239d6eabd595db56852535992c898eba72",
        "type": "github"
      },
      "original": {
@@ -41,11 +41,11 @@
    },
    "nixpkgs": {
      "locked": {
-        "lastModified": 1745234285,
-        "narHash": "sha256-GfpyMzxwkfgRVN0cTGQSkTC0OHhEkv3Jf6Tcjm//qZ0=",
+        "lastModified": 1765472234,
+        "narHash": "sha256-9VvC20PJPsleGMewwcWYKGzDIyjckEz8uWmT0vCDYK0=",
        "owner": "nixos",
        "repo": "nixpkgs",
-        "rev": "c11863f1e964833214b767f4a369c6e6a7aba141",
+        "rev": "2fbfb1d73d239d2402a8fe03963e37aab15abe8b",
        "type": "github"
      },
      "original": {
@@ -100,11 +100,11 @@
        ]
      },
      "locked": {
-        "lastModified": 1744961264,
-        "narHash": "sha256-aRmUh0AMwcbdjJHnytg1e5h5ECcaWtIFQa6d9gI85AI=",
+        "lastModified": 1762938485,
+        "narHash": "sha256-AlEObg0syDl+Spi4LsZIBrjw+snSVU4T8MOeuZJUJjM=",
        "owner": "numtide",
        "repo": "treefmt-nix",
-        "rev": "8d404a69efe76146368885110f29a2ca3700bee6",
+        "rev": "5b4ee75aeefd1e2d5a1cc43cf6ba65eba75e83e4",
        "type": "github"
      },
      "original": {
--- a/go.mod
+++ b/go.mod
@@ -3,7 +3,7 @@ module github.com/danielmiessler/fabric
 go 1.25.1

 require (
-	github.com/anthropics/anthropic-sdk-go v1.13.0
+	github.com/anthropics/anthropic-sdk-go v1.19.0
 	github.com/atotto/clipboard v0.1.4
 	github.com/aws/aws-sdk-go-v2 v1.39.0
 	github.com/aws/aws-sdk-go-v2/config v1.31.8
@@ -21,7 +21,7 @@ require (
 	github.com/mattn/go-sqlite3 v1.14.28
 	github.com/nicksnyder/go-i18n/v2 v2.6.0
 	github.com/ollama/ollama v0.11.7
-	github.com/openai/openai-go v1.8.2
+	github.com/openai/openai-go v1.12.0
 	github.com/otiai10/copy v1.14.1
 	github.com/pkg/errors v0.9.1
 	github.com/samber/lo v1.50.0
@@ -29,7 +29,7 @@ require (
 	github.com/spf13/cobra v1.9.1
 	github.com/stretchr/testify v1.11.1
 	golang.org/x/oauth2 v0.30.0
-	golang.org/x/text v0.28.0
+	golang.org/x/text v0.31.0
 	google.golang.org/api v0.247.0
 	gopkg.in/yaml.v3 v3.0.1
 )
@@ -118,11 +118,11 @@ require (
 	go.opentelemetry.io/otel/metric v1.36.0 // indirect
 	go.opentelemetry.io/otel/trace v1.36.0 // indirect
 	golang.org/x/arch v0.18.0 // indirect
-	golang.org/x/crypto v0.41.0 // indirect
+	golang.org/x/crypto v0.45.0 // indirect
 	golang.org/x/exp v0.0.0-20250531010427-b6e5de432a8b // indirect
-	golang.org/x/net v0.43.0 // indirect
-	golang.org/x/sync v0.16.0 // indirect
-	golang.org/x/sys v0.35.0 // indirect
+	golang.org/x/net v0.47.0 // indirect
+	golang.org/x/sync v0.18.0 // indirect
+	golang.org/x/sys v0.38.0 // indirect
 	google.golang.org/genai v1.17.0
 	google.golang.org/genproto/googleapis/api v0.0.0-20250818200422-3122310a409c // indirect
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20250818200422-3122310a409c // indirect
--- a/go.sum
+++ b/go.sum
@@ -27,8 +27,10 @@ github.com/andybalholm/cascadia v1.3.3 h1:AG2YHrzJIm4BZ19iwJ/DAua6Btl3IwJX+VI4kk
 github.com/andybalholm/cascadia v1.3.3/go.mod h1:xNd9bqTn98Ln4DwST8/nG+H0yuB8Hmgu1YHNnWw0GeA=
 github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be h1:9AeTilPcZAjCFIImctFaOjnTIavg87rW78vTPkQqLI8=
 github.com/anmitsu/go-shlex v0.0.0-20200514113438-38f4b401e2be/go.mod h1:ySMOLuWl6zY27l47sB3qLNK6tF2fkHG55UZxx8oIVo4=
-github.com/anthropics/anthropic-sdk-go v1.13.0 h1:Bhbe8sRoDPtipttg8bQYrMCKe2b79+q6rFW1vOKEUKI=
-github.com/anthropics/anthropic-sdk-go v1.13.0/go.mod h1:WTz31rIUHUHqai2UslPpw5CwXrQP3geYBioRV4WOLvE=
+github.com/anthropics/anthropic-sdk-go v1.16.0 h1:nRkOFDqYXsHteoIhjdJr/5dsiKbFF3rflSv8ax50y8o=
+github.com/anthropics/anthropic-sdk-go v1.16.0/go.mod h1:WTz31rIUHUHqai2UslPpw5CwXrQP3geYBioRV4WOLvE=
+github.com/anthropics/anthropic-sdk-go v1.19.0 h1:mO6E+ffSzLRvR/YUH9KJC0uGw0uV8GjISIuzem//3KE=
+github.com/anthropics/anthropic-sdk-go v1.19.0/go.mod h1:WTz31rIUHUHqai2UslPpw5CwXrQP3geYBioRV4WOLvE=
 github.com/araddon/dateparse v0.0.0-20210429162001-6b43995a97de h1:FxWPpzIjnTlhPwqqXc4/vE0f7GvRjuAsbW+HOIe8KnA=
 github.com/araddon/dateparse v0.0.0-20210429162001-6b43995a97de/go.mod h1:DCaWoUhZrYW9p1lxo/cm8EmUOOzAPSEZNGF2DK1dJgw=
 github.com/armon/go-socks5 v0.0.0-20160902184237-e75332964ef5 h1:0CwZNZbxp69SHPdPJAN/hZIm0C4OItdklCFmMRWYpio=
@@ -201,8 +203,8 @@ github.com/ollama/ollama v0.11.7 h1:CuYjaJ/YEnvLDpJocJbbVdpdVFyGA/OP6lKFyzZD4dI=
 github.com/ollama/ollama v0.11.7/go.mod h1:9+1//yWPsDE2u+l1a5mpaKrYw4VdnSsRU3ioq5BvMms=
 github.com/onsi/gomega v1.34.1 h1:EUMJIKUjM8sKjYbtxQI9A4z2o+rruxnzNvpknOXie6k=
 github.com/onsi/gomega v1.34.1/go.mod h1:kU1QgUvBDLXBJq618Xvm2LUX6rSAfRaFRTcdOeDLwwY=
-github.com/openai/openai-go v1.8.2 h1:UqSkJ1vCOPUpz9Ka5tS0324EJFEuOvMc+lA/EarJWP8=
-github.com/openai/openai-go v1.8.2/go.mod h1:g461MYGXEXBVdV5SaR/5tNzNbSfwTBBefwc+LlDCK0Y=
+github.com/openai/openai-go v1.12.0 h1:NBQCnXzqOTv5wsgNC36PrFEiskGfO5wccfCWDo9S1U0=
+github.com/openai/openai-go v1.12.0/go.mod h1:g461MYGXEXBVdV5SaR/5tNzNbSfwTBBefwc+LlDCK0Y=
 github.com/otiai10/copy v1.14.1 h1:5/7E6qsUMBaH5AnQ0sSLzzTg1oTECmcCmT6lvF45Na8=
 github.com/otiai10/copy v1.14.1/go.mod h1:oQwrEDDOci3IM8dJF0d8+jnbfPDllW6vUjNc3DoZm9I=
 github.com/otiai10/mint v1.6.3 h1:87qsV/aw1F5as1eH1zS/yqHY85ANKVMgkDrf9rcxbQs=
@@ -290,6 +292,8 @@ golang.org/x/crypto v0.23.0/go.mod h1:CKFgDieR+mRhux2Lsu27y0fO304Db0wZe70UKqHu0v
 golang.org/x/crypto v0.31.0/go.mod h1:kDsLvtWBEx7MV9tJOj9bnXsPbxwJQ6csT/x4KIN4Ssk=
 golang.org/x/crypto v0.41.0 h1:WKYxWedPGCTVVl5+WHSSrOBT0O8lx32+zxmHxijgXp4=
 golang.org/x/crypto v0.41.0/go.mod h1:pO5AFd7FA68rFak7rOAGVuygIISepHftHnr8dr6+sUc=
+golang.org/x/crypto v0.45.0 h1:jMBrvKuj23MTlT0bQEOBcAE0mjg8mK9RXFhRH6nyF3Q=
+golang.org/x/crypto v0.45.0/go.mod h1:XTGrrkGJve7CYK7J8PEww4aY7gM3qMCElcJQ8n8JdX4=
 golang.org/x/exp v0.0.0-20250531010427-b6e5de432a8b h1:QoALfVG9rhQ/M7vYDScfPdWjGL9dlsVVM5VGh7aKoAA=
 golang.org/x/exp v0.0.0-20250531010427-b6e5de432a8b/go.mod h1:U6Lno4MTRCDY+Ba7aCcauB9T60gsv5s4ralQzP72ZoQ=
 golang.org/x/mod v0.6.0-dev.0.20220419223038-86c51ed26bb4/go.mod h1:jJ57K6gSWd91VN4djpZkiMVwK6gcyfeH4XE8wZrZaV4=
@@ -309,6 +313,8 @@ golang.org/x/net v0.25.0/go.mod h1:JkAGAh7GEvH74S6FOH42FLoXpXbE/aqXSrIQjXgsiwM=
 golang.org/x/net v0.33.0/go.mod h1:HXLR5J+9DxmrqMwG9qjGCxZ+zKXxBru04zlTvWlWuN4=
 golang.org/x/net v0.43.0 h1:lat02VYK2j4aLzMzecihNvTlJNQUq316m2Mr9rnM6YE=
 golang.org/x/net v0.43.0/go.mod h1:vhO1fvI4dGsIjh73sWfUVjj3N7CA9WkKJNQm2svM6Jg=
+golang.org/x/net v0.47.0 h1:Mx+4dIFzqraBXUugkia1OOvlD6LemFo1ALMHjrXDOhY=
+golang.org/x/net v0.47.0/go.mod h1:/jNxtkgq5yWUGYkaZGqo27cfGZ1c5Nen03aYrrKpVRU=
 golang.org/x/oauth2 v0.30.0 h1:dnDm7JmhM45NNpd8FDDeLhK6FwqbOf4MLCM9zb1BOHI=
 golang.org/x/oauth2 v0.30.0/go.mod h1:B++QgG3ZKulg6sRPGD/mqlHQs5rB3Ml9erfeDY7xKlU=
 golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
@@ -320,6 +326,8 @@ golang.org/x/sync v0.7.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
 golang.org/x/sync v0.10.0/go.mod h1:Czt+wKu1gCyEFDUtn0jG5QVvpJ6rzVqr5aXyt9drQfk=
 golang.org/x/sync v0.16.0 h1:ycBJEhp9p4vXvUZNszeOq0kGTPghopOL8q0fq3vstxw=
 golang.org/x/sync v0.16.0/go.mod h1:1dzgHSNfp02xaA81J2MS99Qcpr2w7fw1gpm99rleRqA=
+golang.org/x/sync v0.18.0 h1:kr88TuHDroi+UVf+0hZnirlk8o8T+4MrK6mr60WkH/I=
+golang.org/x/sync v0.18.0/go.mod h1:9KTHXmSnoGruLpwFjVSX0lNNA75CykiMECbovNTZqGI=
 golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY=
 golang.org/x/sys v0.0.0-20191026070338-33540a1f6037/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20201119102817-f84b799fce68/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
@@ -338,6 +346,8 @@ golang.org/x/sys v0.20.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
 golang.org/x/sys v0.28.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
 golang.org/x/sys v0.35.0 h1:vz1N37gP5bs89s7He8XuIYXpyY0+QlsKmzipCbUtyxI=
 golang.org/x/sys v0.35.0/go.mod h1:BJP2sWEmIv4KK5OTEluFJCKSidICx8ciO85XgH3Ak8k=
+golang.org/x/sys v0.38.0 h1:3yZWxaJjBmCWXqhN1qh02AkOnCQ1poK6oF+a7xWL6Gc=
+golang.org/x/sys v0.38.0/go.mod h1:OgkHotnGiDImocRcuBABYBEXf8A9a87e/uXjp9XT3ks=
 golang.org/x/telemetry v0.0.0-20240228155512-f48c80bd79b2/go.mod h1:TeRTkGYfJXctD9OcfyVLyj2J3IxLnKwHJR8f4D8a3YE=
 golang.org/x/term v0.0.0-20201126162022-7de9c90e9dd1/go.mod h1:bj7SfCRtBDWHUb9snDiAeCFNEtKQo2Wmx5Cou7ajbmo=
 golang.org/x/term v0.0.0-20210927222741-03fcf44c2211/go.mod h1:jbD1KX2456YbFQfuXm/mYQcufACuNUgVhRMnK/tPxf8=
@@ -349,6 +359,7 @@ golang.org/x/term v0.20.0/go.mod h1:8UkIAJTvZgivsXaD6/pH6U9ecQzZ45awqEOzuCvwpFY=
 golang.org/x/term v0.27.0/go.mod h1:iMsnZpn0cago0GOrHO2+Y7u7JPn5AylBrcoWkElMTSM=
 golang.org/x/term v0.34.0 h1:O/2T7POpk0ZZ7MAzMeWFSg6S5IpWd/RXDlM9hgM3DR4=
 golang.org/x/term v0.34.0/go.mod h1:5jC53AEywhIVebHgPVeg0mj8OD3VO9OzclacVrqpaAw=
+golang.org/x/term v0.37.0 h1:8EGAD0qCmHYZg6J17DvsMy9/wJ7/D/4pV/wfnld5lTU=
 golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.6/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
@@ -361,6 +372,8 @@ golang.org/x/text v0.15.0/go.mod h1:18ZOQIKpY8NJVqYksKHtTdi31H5itFRjB5/qKTNYzSU=
 golang.org/x/text v0.21.0/go.mod h1:4IBbMaMmOPCJ8SecivzSH54+73PCFmPWxNTLm+vZkEQ=
 golang.org/x/text v0.28.0 h1:rhazDwis8INMIwQ4tpjLDzUhx6RlXqZNPEM0huQojng=
 golang.org/x/text v0.28.0/go.mod h1:U8nCwOR8jO/marOQ0QbDiOngZVEBB7MAiitBuMjXiNU=
+golang.org/x/text v0.31.0 h1:aC8ghyu4JhP8VojJ2lEHBnochRno1sgL6nEi9WGFGMM=
+golang.org/x/text v0.31.0/go.mod h1:tKRAlv61yKIjGGHX/4tP1LTbc13YSec1pxVEWXzfoeM=
 golang.org/x/tools v0.0.0-20180917221912-90fa682c2a6e/go.mod h1:n7NCudcB/nEzxVGmLbDWY5pfWTLqBcC2KZ6jyYvM4mQ=
 golang.org/x/tools v0.0.0-20191119224855-298f0cb1881e/go.mod h1:b+2E5dAYhXwXZwtnZ6UAqBI28+e2cm9otk0dWdXHAEo=
 golang.org/x/tools v0.1.12/go.mod h1:hNGJHUnrk76NpqgfD5Aqm5Crs+Hm0VOH/i9J2+nxYbc=
--- a/internal/cli/flags.go
+++ b/internal/cli/flags.go
@@ -8,6 +8,7 @@ import (
 	"os"
 	"path/filepath"
 	"reflect"
+	"slices"
 	"strconv"
 	"strings"

@@ -35,7 +36,7 @@ type Flags struct {
 	TopP                            float64              `short:"T" long:"topp" yaml:"topp" description:"Set top P" default:"0.9"`
 	Stream                          bool                 `short:"s" long:"stream" yaml:"stream" description:"Stream"`
 	PresencePenalty                 float64              `short:"P" long:"presencepenalty" yaml:"presencepenalty" description:"Set presence penalty" default:"0.0"`
-	Raw                             bool                 `short:"r" long:"raw" yaml:"raw" description:"Use the defaults of the model without sending chat options (like temperature etc.) and use the user role instead of the system role for patterns."`
+	Raw                             bool                 `short:"r" long:"raw" yaml:"raw" description:"Use the defaults of the model without sending chat options (temperature, top_p, etc.). Only affects OpenAI-compatible providers. Anthropic models always use smart parameter selection to comply with model-specific requirements."`
 	FrequencyPenalty                float64              `short:"F" long:"frequencypenalty" yaml:"frequencypenalty" description:"Set frequency penalty" default:"0.0"`
 	ListPatterns                    bool                 `short:"l" long:"listpatterns" description:"List all patterns"`
 	ListAllModels                   bool                 `short:"L" long:"listmodels" description:"List all available models"`
@@ -115,7 +116,7 @@ func Init() (ret *Flags, err error) {

 	// Create mapping from flag names (both short and long) to yaml tag names
 	flagToYamlTag := make(map[string]string)
-	t := reflect.TypeOf(Flags{})
+	t := reflect.TypeFor[Flags]()
 	for i := 0; i < t.NumField(); i++ {
 		field := t.Field(i)
 		yamlTag := field.Tag.Get("yaml")
@@ -224,14 +225,14 @@ func Init() (ret *Flags, err error) {
 }

 func parseDebugLevel(args []string) int {
-	for i := 0; i < len(args); i++ {
+	for i := range args {
 		arg := args[i]
 		if arg == "--debug" && i+1 < len(args) {
 			if lvl, err := strconv.Atoi(args[i+1]); err == nil {
 				return lvl
 			}
-		} else if strings.HasPrefix(arg, "--debug=") {
-			if lvl, err := strconv.Atoi(strings.TrimPrefix(arg, "--debug=")); err == nil {
+		} else if after, ok := strings.CutPrefix(arg, "--debug="); ok {
+			if lvl, err := strconv.Atoi(after); err == nil {
 				return lvl
 			}
 		}
@@ -241,8 +242,8 @@ func parseDebugLevel(args []string) int {

 func extractFlag(arg string) string {
 	var flag string
-	if strings.HasPrefix(arg, "--") {
-		flag = strings.TrimPrefix(arg, "--")
+	if after, ok := strings.CutPrefix(arg, "--"); ok {
+		flag = after
 		if i := strings.Index(flag, "="); i > 0 {
 			flag = flag[:i]
 		}
@@ -348,10 +349,8 @@ func validateImageFile(imagePath string) error {
 	ext := strings.ToLower(filepath.Ext(imagePath))
 	validExtensions := []string{".png", ".jpeg", ".jpg", ".webp"}

-	for _, validExt := range validExtensions {
-		if ext == validExt {
-			return nil // Valid extension found
-		}
+	if slices.Contains(validExtensions, ext) {
+		return nil // Valid extension found
 	}

 	return fmt.Errorf("%s", fmt.Sprintf(i18n.T("invalid_image_file_extension"), ext))
@@ -370,13 +369,7 @@ func validateImageParameters(imagePath, size, quality, background string, compre
 	// Validate size
 	if size != "" {
 		validSizes := []string{"1024x1024", "1536x1024", "1024x1536", "auto"}
-		valid := false
-		for _, validSize := range validSizes {
-			if size == validSize {
-				valid = true
-				break
-			}
-		}
+		valid := slices.Contains(validSizes, size)
 		if !valid {
 			return fmt.Errorf("%s", fmt.Sprintf(i18n.T("invalid_image_size"), size))
 		}
@@ -385,13 +378,7 @@ func validateImageParameters(imagePath, size, quality, background string, compre
 	// Validate quality
 	if quality != "" {
 		validQualities := []string{"low", "medium", "high", "auto"}
-		valid := false
-		for _, validQuality := range validQualities {
-			if quality == validQuality {
-				valid = true
-				break
-			}
-		}
+		valid := slices.Contains(validQualities, quality)
 		if !valid {
 			return fmt.Errorf("%s", fmt.Sprintf(i18n.T("invalid_image_quality"), quality))
 		}
@@ -400,13 +387,7 @@ func validateImageParameters(imagePath, size, quality, background string, compre
 	// Validate background
 	if background != "" {
 		validBackgrounds := []string{"opaque", "transparent"}
-		valid := false
-		for _, validBackground := range validBackgrounds {
-			if background == validBackground {
-				valid = true
-				break
-			}
-		}
+		valid := slices.Contains(validBackgrounds, background)
 		if !valid {
 			return fmt.Errorf("%s", fmt.Sprintf(i18n.T("invalid_image_background"), background))
 		}
--- a/internal/cli/help.go
+++ b/internal/cli/help.go
@@ -137,8 +137,7 @@ func (h *TranslatedHelpWriter) getTranslatedDescription(flagName string) string

 // getOriginalDescription retrieves the original description from struct tags
 func (h *TranslatedHelpWriter) getOriginalDescription(flagName string) string {
-	flags := &Flags{}
-	flagsType := reflect.TypeOf(flags).Elem()
+	flagsType := reflect.TypeFor[Flags]()

 	for i := 0; i < flagsType.NumField(); i++ {
 		field := flagsType.Field(i)
@@ -184,10 +183,10 @@ func detectLanguageFromArgs() string {
 			if i+1 < len(args) {
 				return args[i+1]
 			}
-		} else if strings.HasPrefix(arg, "--language=") {
-			return strings.TrimPrefix(arg, "--language=")
-		} else if strings.HasPrefix(arg, "-g=") {
-			return strings.TrimPrefix(arg, "-g=")
+		} else if after, ok := strings.CutPrefix(arg, "--language="); ok {
+			return after
+		} else if after, ok := strings.CutPrefix(arg, "-g="); ok {
+			return after
 		} else if runtime.GOOS == "windows" && strings.HasPrefix(arg, "/g:") {
 			return strings.TrimPrefix(arg, "/g:")
 		} else if runtime.GOOS == "windows" && strings.HasPrefix(arg, "/g=") {
@@ -218,8 +217,7 @@ func detectLanguageFromEnv() string {
 // writeAllFlags writes all flags with translated descriptions
 func (h *TranslatedHelpWriter) writeAllFlags() {
 	// Use direct reflection on the Flags struct to get all flag definitions
-	flags := &Flags{}
-	flagsType := reflect.TypeOf(flags).Elem()
+	flagsType := reflect.TypeFor[Flags]()

 	for i := 0; i < flagsType.NumField(); i++ {
 		field := flagsType.Field(i)
@@ -274,10 +272,7 @@ func (h *TranslatedHelpWriter) writeAllFlags() {

 		// Pad to align descriptions
 		flagStr := flagLine.String()
-		padding := 34 - len(flagStr)
-		if padding < 2 {
-			padding = 2
-		}
+		padding := max(34-len(flagStr), 2)

 		fmt.Fprintf(h.writer, "%s%s%s", flagStr, strings.Repeat(" ", padding), description)

--- a/internal/cli/listing.go
+++ b/internal/cli/listing.go
@@ -39,6 +39,11 @@ func handleListingCommands(currentFlags *Flags, fabricDb *fsdb.Db, registry *cor
 		if models, err = registry.VendorManager.GetModels(); err != nil {
 			return true, err
 		}
+
+		if currentFlags.Vendor != "" {
+			models = models.FilterByVendor(currentFlags.Vendor)
+		}
+
 		if currentFlags.ShellCompleteOutput {
 			models.Print(true)
 		} else {
--- a/internal/cli/output.go
+++ b/internal/cli/output.go
@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
+	"slices"
 	"strings"

 	"github.com/atotto/clipboard"
@@ -29,6 +30,9 @@ func CreateOutputFile(message string, fileName string) (err error) {
 		return
 	}
 	defer file.Close()
+	if !strings.HasSuffix(message, "\n") {
+		message += "\n"
+	}
 	if _, err = file.WriteString(message); err != nil {
 		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("error_writing_to_file"), err))
 	} else {
@@ -63,10 +67,5 @@ func CreateAudioOutputFile(audioData []byte, fileName string) (err error) {
 func IsAudioFormat(fileName string) bool {
 	ext := strings.ToLower(filepath.Ext(fileName))
 	audioExts := []string{".wav", ".mp3", ".m4a", ".aac", ".ogg", ".flac"}
-	for _, audioExt := range audioExts {
-		if ext == audioExt {
-			return true
-		}
-	}
-	return false
+	return slices.Contains(audioExts, ext)
 }
--- a/internal/cli/output_test.go
+++ b/internal/cli/output_test.go
@@ -24,5 +24,34 @@ func TestCreateOutputFile(t *testing.T) {
 		t.Fatalf("CreateOutputFile() error = %v", err)
 	}

-	defer os.Remove(fileName)
+	t.Cleanup(func() { os.Remove(fileName) })
+
+	data, err := os.ReadFile(fileName)
+	if err != nil {
+		t.Fatalf("failed to read output file: %v", err)
+	}
+
+	expected := message + "\n"
+	if string(data) != expected {
+		t.Fatalf("expected file contents %q, got %q", expected, data)
+	}
+}
+
+func TestCreateOutputFileMessageWithTrailingNewline(t *testing.T) {
+	fileName := "test_output_with_newline.txt"
+	message := "test message with newline\n"
+
+	if err := CreateOutputFile(message, fileName); err != nil {
+		t.Fatalf("CreateOutputFile() error = %v", err)
+	}
+	t.Cleanup(func() { os.Remove(fileName) })
+
+	data, err := os.ReadFile(fileName)
+	if err != nil {
+		t.Fatalf("failed to read output file: %v", err)
+	}
+
+	if string(data) != message {
+		t.Fatalf("expected file contents %q, got %q", message, data)
+	}
 }
--- a/internal/cli/transcribe.go
+++ b/internal/cli/transcribe.go
@@ -17,8 +17,9 @@ func handleTranscription(flags *Flags, registry *core.PluginRegistry) (message s
 	if vendorName == "" {
 		vendorName = "OpenAI"
 	}
-	vendor, ok := registry.VendorManager.VendorsByName[vendorName]
-	if !ok {
+
+	vendor := registry.VendorManager.FindByName(vendorName)
+	if vendor == nil {
 		return "", fmt.Errorf("%s", fmt.Sprintf(i18n.T("vendor_not_configured"), vendorName))
 	}
 	tr, ok := vendor.(transcriber)
--- a/internal/core/chatter.go
+++ b/internal/core/chatter.go
@@ -32,11 +32,9 @@ type Chatter struct {

 // Send processes a chat request and applies file changes for create_coding_feature pattern
 func (o *Chatter) Send(request *domain.ChatRequest, opts *domain.ChatOptions) (session *fsdb.Session, err error) {
-	modelToUse := opts.Model
-	if modelToUse == "" {
-		modelToUse = o.model
-	}
-	if o.vendor.NeedsRawMode(modelToUse) {
+	// Use o.model (normalized) for NeedsRawMode check instead of opts.Model
+	// This ensures case-insensitive model names work correctly (e.g., "GPT-5" → "gpt-5")
+	if o.vendor.NeedsRawMode(o.model) {
 		opts.Raw = true
 	}
 	if session, err = o.BuildSession(request, opts.Raw); err != nil {
@@ -57,6 +55,10 @@ func (o *Chatter) Send(request *domain.ChatRequest, opts *domain.ChatOptions) (s

 	if opts.Model == "" {
 		opts.Model = o.model
+	} else {
+		// Ensure opts.Model uses the normalized name from o.model if they refer to the same model
+		// This handles cases where user provides "GPT-5" but we've normalized it to "gpt-5"
+		opts.Model = o.model
 	}

 	if opts.ModelContextLength == 0 {
@@ -69,6 +71,7 @@ func (o *Chatter) Send(request *domain.ChatRequest, opts *domain.ChatOptions) (s
 		responseChan := make(chan string)
 		errChan := make(chan error, 1)
 		done := make(chan struct{})
+		printedStream := false

 		go func() {
 			defer close(done)
@@ -81,9 +84,14 @@ func (o *Chatter) Send(request *domain.ChatRequest, opts *domain.ChatOptions) (s
 			message += response
 			if !opts.SuppressThink {
 				fmt.Print(response)
+				printedStream = true
 			}
 		}

+		if printedStream && !opts.SuppressThink && !strings.HasSuffix(message, "\n") {
+			fmt.Println()
+		}
+
 		// Wait for goroutine to finish
 		<-done

@@ -175,7 +183,7 @@ func (o *Chatter) BuildSession(request *domain.ChatRequest, raw bool) (session *
 	if request.Message == nil {
 		request.Message = &chat.ChatCompletionMessage{
 			Role:    chat.ChatMessageRoleUser,
-			Content: " ",
+			Content: "",
 		}
 	}

--- a/internal/core/plugin_registry.go
+++ b/internal/core/plugin_registry.go
@@ -222,9 +222,8 @@ func (o *PluginRegistry) Setup() (err error) {
 				}
 			}

-			if _, ok := o.VendorManager.VendorsByName[plugin.GetName()]; !ok {
-				var vendor ai.Vendor
-				if vendor, ok = plugin.(ai.Vendor); ok {
+			if o.VendorManager.FindByName(plugin.GetName()) == nil {
+				if vendor, ok := plugin.(ai.Vendor); ok {
 					o.VendorManager.AddVendors(vendor)
 				}
 			}
@@ -330,11 +329,22 @@ func (o *PluginRegistry) GetChatter(model string, modelContextLength int, vendor
 		if models, err = vendorManager.GetModels(); err != nil {
 			return
 		}
+
+		// Normalize model name to match actual available model (case-insensitive)
+		// This must be done BEFORE checking vendor availability
+		actualModelName := models.FindModelNameCaseInsensitive(model)
+		if actualModelName != "" {
+			model = actualModelName // Use normalized name for all subsequent checks
+		}
+
 		if vendorName != "" {
 			// ensure vendor exists and provides model
 			ret.vendor = vendorManager.FindByName(vendorName)
 			availableVendors := models.FindGroupsByItem(model)
-			if ret.vendor == nil || !lo.Contains(availableVendors, vendorName) {
+			vendorAvailable := lo.ContainsBy(availableVendors, func(name string) bool {
+				return strings.EqualFold(name, vendorName)
+			})
+			if ret.vendor == nil || !vendorAvailable {
 				err = fmt.Errorf("model %s not available for vendor %s", model, vendorName)
 				return
 			}
@@ -345,6 +355,7 @@ func (o *PluginRegistry) GetChatter(model string, modelContextLength int, vendor
 			}
 			ret.vendor = vendorManager.FindByName(models.FindGroupsByItemFirst(model))
 		}
+
 		ret.model = model
 	}

--- a/internal/domain/file_manager.go
+++ b/internal/domain/file_manager.go
@@ -5,6 +5,7 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
+	"slices"
 	"strings"
 )

@@ -146,14 +147,7 @@ func fixInvalidEscapes(jsonStr string) string {
 		// Check for escape sequences only inside strings
 		if inQuotes && ch == '\\' && i+1 < len(jsonStr) {
 			nextChar := jsonStr[i+1]
-			isValid := false
-
-			for _, validEscape := range validEscapes {
-				if nextChar == validEscape {
-					isValid = true
-					break
-				}
-			}
+			isValid := slices.Contains(validEscapes, nextChar)

 			if !isValid {
 				// Invalid escape sequence - add an extra backslash
--- a/internal/i18n/locales/de.json
+++ b/internal/i18n/locales/de.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "Anbieter %s unterstützt keine Audio-Transkription",
  "transcription_model_required": "Transkriptionsmodell ist erforderlich (verwende --transcribe-model)",
  "youtube_not_configured": "YouTube ist nicht konfiguriert, bitte führe das Setup-Verfahren aus",
+  "youtube_api_key_required": "YouTube API-Schlüssel für Kommentare und Metadaten erforderlich. Führe 'fabric --setup' aus, um zu konfigurieren",
+  "youtube_ytdlp_not_found": "yt-dlp wurde nicht in PATH gefunden. Bitte installiere yt-dlp, um die YouTube-Transkript-Funktionalität zu nutzen",
+  "youtube_invalid_url": "ungültige YouTube-URL, kann keine Video- oder Playlist-ID abrufen: '%s'",
+  "youtube_url_is_playlist_not_video": "URL ist eine Playlist, kein Video",
+  "youtube_no_video_id_found": "keine Video-ID in URL gefunden",
+  "youtube_rate_limit_exceeded": "YouTube-Ratenlimit überschritten. Versuche es später erneut oder verwende andere yt-dlp-Argumente wie '--sleep-requests 1', um Anfragen zu verlangsamen.",
+  "youtube_auth_required_bot_detection": "YouTube erfordert Authentifizierung (Bot-Erkennung). Verwende --yt-dlp-args='--cookies-from-browser BROWSER' wobei BROWSER chrome, firefox, brave usw. sein kann.",
+  "youtube_ytdlp_stderr_error": "Fehler beim Lesen von yt-dlp stderr",
+  "youtube_invalid_ytdlp_arguments": "ungültige yt-dlp-Argumente: %v",
+  "youtube_failed_create_temp_dir": "temporäres Verzeichnis konnte nicht erstellt werden: %v",
+  "youtube_no_transcript_content": "kein Transkriptinhalt in VTT-Datei gefunden",
+  "youtube_no_vtt_files_found": "keine VTT-Dateien im Verzeichnis gefunden",
+  "youtube_failed_walk_directory": "Verzeichnis konnte nicht durchlaufen werden: %v",
+  "youtube_error_getting_video_details": "Fehler beim Abrufen der Videodetails: %v",
+  "youtube_invalid_duration_string": "ungültige Dauer-Zeichenfolge: %s",
+  "youtube_error_getting_metadata": "Fehler beim Abrufen der Video-Metadaten: %v",
+  "youtube_error_parsing_duration": "Fehler beim Parsen der Videodauer: %v",
+  "youtube_error_getting_comments": "Fehler beim Abrufen der Kommentare: %v",
+  "youtube_error_saving_csv": "Fehler beim Speichern der Videos in CSV: %v",
+  "youtube_no_video_found_with_id": "kein Video mit ID gefunden: %s",
+  "youtube_invalid_timestamp_format": "ungültiges Zeitstempel-Format: %s",
+  "youtube_empty_seconds_string": "leere Sekunden-Zeichenfolge",
+  "youtube_invalid_seconds_format": "ungültiges Sekundenformat %q: %w",
  "error_fetching_playlist_videos": "Fehler beim Abrufen der Playlist-Videos: %w",
+  "openai_api_base_url_not_configured": "API-Basis-URL für Anbieter %s nicht konfiguriert",
+  "openai_failed_to_create_models_url": "Modell-URL konnte nicht erstellt werden: %w",
+  "openai_unexpected_status_code_with_body": "unerwarteter Statuscode: %d von Anbieter %s, Antwort: %s",
+  "openai_unexpected_status_code_read_error_partial": "unerwarteter Statuscode: %d von Anbieter %s (Fehler beim Lesen: %v), teilweise Antwort: %s",
+  "openai_unexpected_status_code_read_error": "unerwarteter Statuscode: %d von Anbieter %s (Fehler beim Lesen der Antwort: %v)",
+  "openai_unable_to_parse_models_response": "Modell-Antwort konnte nicht geparst werden; rohe Antwort: %s",
  "scraping_not_configured": "Scraping-Funktionalität ist nicht konfiguriert. Bitte richte Jina ein, um Scraping zu aktivieren",
  "could_not_determine_home_dir": "konnte Benutzer-Home-Verzeichnis nicht bestimmen: %w",
  "could_not_stat_env_file": "konnte .env-Datei nicht überprüfen: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Top P festlegen",
  "stream_help": "Streaming",
  "set_presence_penalty": "Präsenzstrafe festlegen",
-  "use_model_defaults_raw_help": "Verwende die Standardwerte des Modells ohne Senden von Chat-Optionen (wie Temperatur usw.) und verwende die Benutzerrolle anstelle der Systemrolle für Muster.",
+  "use_model_defaults_raw_help": "Verwende die Standardwerte des Modells, ohne Chat-Optionen (temperature, top_p usw.) zu senden. Gilt nur für OpenAI-kompatible Anbieter. Anthropic-Modelle verwenden stets eine intelligente Parameterauswahl, um modell-spezifische Anforderungen einzuhalten.",
  "set_frequency_penalty": "Häufigkeitsstrafe festlegen",
  "list_all_patterns": "Alle Muster auflisten",
  "list_all_available_models": "Alle verfügbaren Modelle auflisten",
--- a/internal/i18n/locales/en.json
+++ b/internal/i18n/locales/en.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "vendor %s does not support audio transcription",
  "transcription_model_required": "transcription model is required (use --transcribe-model)",
  "youtube_not_configured": "YouTube is not configured, please run the setup procedure",
+  "youtube_api_key_required": "YouTube API key required for comments and metadata. Run 'fabric --setup' to configure",
+  "youtube_ytdlp_not_found": "yt-dlp not found in PATH. Please install yt-dlp to use YouTube transcript functionality",
+  "youtube_invalid_url": "invalid YouTube URL, can't get video or playlist ID: '%s'",
+  "youtube_url_is_playlist_not_video": "URL is a playlist, not a video",
+  "youtube_no_video_id_found": "no video ID found in URL",
+  "youtube_rate_limit_exceeded": "YouTube rate limit exceeded. Try again later or use different yt-dlp arguments like '--sleep-requests 1' to slow down requests.",
+  "youtube_auth_required_bot_detection": "YouTube requires authentication (bot detection). Use --yt-dlp-args='--cookies-from-browser BROWSER' where BROWSER is chrome, firefox, brave, etc.",
+  "youtube_ytdlp_stderr_error": "Error reading yt-dlp stderr",
+  "youtube_invalid_ytdlp_arguments": "invalid yt-dlp arguments: %v",
+  "youtube_failed_create_temp_dir": "failed to create temp directory: %v",
+  "youtube_no_transcript_content": "no transcript content found in VTT file",
+  "youtube_no_vtt_files_found": "no VTT files found in directory",
+  "youtube_failed_walk_directory": "failed to walk directory: %v",
+  "youtube_error_getting_video_details": "error getting video details: %v",
+  "youtube_invalid_duration_string": "invalid duration string: %s",
+  "youtube_error_getting_metadata": "error getting video metadata: %v",
+  "youtube_error_parsing_duration": "error parsing video duration: %v",
+  "youtube_error_getting_comments": "error getting comments: %v",
+  "youtube_error_saving_csv": "error saving videos to CSV: %v",
+  "youtube_no_video_found_with_id": "no video found with ID: %s",
+  "youtube_invalid_timestamp_format": "invalid timestamp format: %s",
+  "youtube_empty_seconds_string": "empty seconds string",
+  "youtube_invalid_seconds_format": "invalid seconds format %q: %w",
  "error_fetching_playlist_videos": "error fetching playlist videos: %w",
+  "openai_api_base_url_not_configured": "API base URL not configured for provider %s",
+  "openai_failed_to_create_models_url": "failed to create models URL: %w",
+  "openai_unexpected_status_code_with_body": "unexpected status code: %d from provider %s, response body: %s",
+  "openai_unexpected_status_code_read_error_partial": "unexpected status code: %d from provider %s (error reading body: %v), partial response: %s",
+  "openai_unexpected_status_code_read_error": "unexpected status code: %d from provider %s (failed to read response body: %v)",
+  "openai_unable_to_parse_models_response": "unable to parse models response; raw response: %s",
  "scraping_not_configured": "scraping functionality is not configured. Please set up Jina to enable scraping",
  "could_not_determine_home_dir": "could not determine user home directory: %w",
  "could_not_stat_env_file": "could not stat .env file: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Set top P",
  "stream_help": "Stream",
  "set_presence_penalty": "Set presence penalty",
-  "use_model_defaults_raw_help": "Use the defaults of the model without sending chat options (like temperature etc.) and use the user role instead of the system role for patterns.",
+  "use_model_defaults_raw_help": "Use the defaults of the model without sending chat options (temperature, top_p, etc.). Only affects OpenAI-compatible providers. Anthropic models always use smart parameter selection to comply with model-specific requirements.",
  "set_frequency_penalty": "Set frequency penalty",
  "list_all_patterns": "List all patterns",
  "list_all_available_models": "List all available models",
--- a/internal/i18n/locales/es.json
+++ b/internal/i18n/locales/es.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "el proveedor %s no admite transcripción de audio",
  "transcription_model_required": "se requiere un modelo de transcripción (usa --transcribe-model)",
  "youtube_not_configured": "YouTube no está configurado, por favor ejecuta el procedimiento de configuración",
+  "youtube_api_key_required": "Se requiere clave de API de YouTube para comentarios y metadatos. Ejecuta 'fabric --setup' para configurar",
+  "youtube_ytdlp_not_found": "yt-dlp no encontrado en PATH. Por favor instala yt-dlp para usar la funcionalidad de transcripción de YouTube",
+  "youtube_invalid_url": "URL de YouTube inválida, no se puede obtener ID de video o lista de reproducción: '%s'",
+  "youtube_url_is_playlist_not_video": "La URL es una lista de reproducción, no un video",
+  "youtube_no_video_id_found": "no se encontró ID de video en la URL",
+  "youtube_rate_limit_exceeded": "Límite de tasa de YouTube excedido. Intenta de nuevo más tarde o usa diferentes argumentos de yt-dlp como '--sleep-requests 1' para ralentizar las solicitudes.",
+  "youtube_auth_required_bot_detection": "YouTube requiere autenticación (detección de bot). Usa --yt-dlp-args='--cookies-from-browser BROWSER' donde BROWSER puede ser chrome, firefox, brave, etc.",
+  "youtube_ytdlp_stderr_error": "Error al leer stderr de yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "argumentos de yt-dlp inválidos: %v",
+  "youtube_failed_create_temp_dir": "falló al crear directorio temporal: %v",
+  "youtube_no_transcript_content": "no se encontró contenido de transcripción en el archivo VTT",
+  "youtube_no_vtt_files_found": "no se encontraron archivos VTT en el directorio",
+  "youtube_failed_walk_directory": "falló al recorrer el directorio: %v",
+  "youtube_error_getting_video_details": "error al obtener detalles del video: %v",
+  "youtube_invalid_duration_string": "cadena de duración inválida: %s",
+  "youtube_error_getting_metadata": "error al obtener metadatos del video: %v",
+  "youtube_error_parsing_duration": "error al analizar la duración del video: %v",
+  "youtube_error_getting_comments": "error al obtener comentarios: %v",
+  "youtube_error_saving_csv": "error al guardar videos en CSV: %v",
+  "youtube_no_video_found_with_id": "no se encontró video con ID: %s",
+  "youtube_invalid_timestamp_format": "formato de marca de tiempo inválido: %s",
+  "youtube_empty_seconds_string": "cadena de segundos vacía",
+  "youtube_invalid_seconds_format": "formato de segundos inválido %q: %w",
  "error_fetching_playlist_videos": "error al obtener videos de la lista de reproducción: %w",
+  "openai_api_base_url_not_configured": "URL base de API no configurada para el proveedor %s",
+  "openai_failed_to_create_models_url": "error al crear URL de modelos: %w",
+  "openai_unexpected_status_code_with_body": "código de estado inesperado: %d del proveedor %s, cuerpo de respuesta: %s",
+  "openai_unexpected_status_code_read_error_partial": "código de estado inesperado: %d del proveedor %s (error al leer cuerpo: %v), respuesta parcial: %s",
+  "openai_unexpected_status_code_read_error": "código de estado inesperado: %d del proveedor %s (error al leer cuerpo de respuesta: %v)",
+  "openai_unable_to_parse_models_response": "no se pudo analizar la respuesta de modelos; respuesta cruda: %s",
  "scraping_not_configured": "la funcionalidad de extracción no está configurada. Por favor configura Jina para habilitar la extracción",
  "could_not_determine_home_dir": "no se pudo determinar el directorio home del usuario: %w",
  "could_not_stat_env_file": "no se pudo verificar el archivo .env: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Establecer top P",
  "stream_help": "Transmitir",
  "set_presence_penalty": "Establecer penalización de presencia",
-  "use_model_defaults_raw_help": "Usar los valores predeterminados del modelo sin enviar opciones de chat (como temperatura, etc.) y usar el rol de usuario en lugar del rol del sistema para patrones.",
+  "use_model_defaults_raw_help": "Utiliza los valores predeterminados del modelo sin enviar opciones de chat (temperature, top_p, etc.). Solo afecta a los proveedores compatibles con OpenAI. Los modelos de Anthropic siempre usan una selección inteligente de parámetros para cumplir los requisitos específicos del modelo.",
  "set_frequency_penalty": "Establecer penalización de frecuencia",
  "list_all_patterns": "Listar todos los patrones",
  "list_all_available_models": "Listar todos los modelos disponibles",
--- a/internal/i18n/locales/fa.json
+++ b/internal/i18n/locales/fa.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "تامین‌کننده %s از رونویسی صوتی پشتیبانی نمی‌کند",
  "transcription_model_required": "مدل رونویسی الزامی است (از --transcribe-model استفاده کنید)",
  "youtube_not_configured": "یوتیوب پیکربندی نشده است، لطفاً روند تنظیمات را اجرا کنید",
+  "youtube_api_key_required": "کلید API یوتیوب برای دریافت نظرات و متادیتا الزامی است. برای پیکربندی 'fabric --setup' را اجرا کنید",
+  "youtube_ytdlp_not_found": "yt-dlp در PATH یافت نشد. لطفاً yt-dlp را نصب کنید تا از قابلیت رونویسی یوتیوب استفاده کنید",
+  "youtube_invalid_url": "URL یوتیوب نامعتبر است، نمی‌توان ID ویدیو یا فهرست پخش را دریافت کرد: '%s'",
+  "youtube_url_is_playlist_not_video": "URL یک فهرست پخش است، نه یک ویدیو",
+  "youtube_no_video_id_found": "هیچ ID ویدیویی در URL یافت نشد",
+  "youtube_rate_limit_exceeded": "محدودیت نرخ یوتیوب فراتر رفته است. بعداً دوباره امتحان کنید یا از آرگومان‌های مختلف yt-dlp مانند '--sleep-requests 1' برای کاهش سرعت درخواست‌ها استفاده کنید.",
+  "youtube_auth_required_bot_detection": "یوتیوب احراز هویت می‌خواهد (تشخیص ربات). از --yt-dlp-args='--cookies-from-browser BROWSER' استفاده کنید که BROWSER می‌تواند chrome، firefox، brave و غیره باشد.",
+  "youtube_ytdlp_stderr_error": "خطا در خواندن stderr yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "آرگومان‌های yt-dlp نامعتبر: %v",
+  "youtube_failed_create_temp_dir": "ایجاد دایرکتوری موقت ناموفق بود: %v",
+  "youtube_no_transcript_content": "محتوای رونوشتی در فایل VTT یافت نشد",
+  "youtube_no_vtt_files_found": "فایل‌های VTT در دایرکتوری یافت نشدند",
+  "youtube_failed_walk_directory": "پیمایش دایرکتوری ناموفق بود: %v",
+  "youtube_error_getting_video_details": "خطا در دریافت جزئیات ویدیو: %v",
+  "youtube_invalid_duration_string": "رشته مدت زمان نامعتبر: %s",
+  "youtube_error_getting_metadata": "خطا در دریافت متادیتای ویدیو: %v",
+  "youtube_error_parsing_duration": "خطا در تجزیه مدت زمان ویدیو: %v",
+  "youtube_error_getting_comments": "خطا در دریافت نظرات: %v",
+  "youtube_error_saving_csv": "خطا در ذخیره ویدیوها در CSV: %v",
+  "youtube_no_video_found_with_id": "هیچ ویدیویی با ID یافت نشد: %s",
+  "youtube_invalid_timestamp_format": "فرمت مهر زمانی نامعتبر: %s",
+  "youtube_empty_seconds_string": "رشته ثانیه خالی",
+  "youtube_invalid_seconds_format": "فرمت ثانیه نامعتبر %q: %w",
  "error_fetching_playlist_videos": "خطا در دریافت ویدیوهای فهرست پخش: %w",
+  "openai_api_base_url_not_configured": "URL پایه API برای ارائه‌دهنده %s پیکربندی نشده است",
+  "openai_failed_to_create_models_url": "ایجاد URL مدل‌ها ناموفق بود: %w",
+  "openai_unexpected_status_code_with_body": "کد وضعیت غیرمنتظره: %d از ارائه‌دهنده %s، پاسخ: %s",
+  "openai_unexpected_status_code_read_error_partial": "کد وضعیت غیرمنتظره: %d از ارائه‌دهنده %s (خطا در خواندن: %v)، پاسخ جزئی: %s",
+  "openai_unexpected_status_code_read_error": "کد وضعیت غیرمنتظره: %d از ارائه‌دهنده %s (خطا در خواندن پاسخ: %v)",
+  "openai_unable_to_parse_models_response": "تجزیه پاسخ مدل‌ها ناموفق بود; پاسخ خام: %s",
  "scraping_not_configured": "قابلیت استخراج داده پیکربندی نشده است. لطفاً Jina را برای فعال‌سازی استخراج تنظیم کنید",
  "could_not_determine_home_dir": "نتوانست دایرکتوری خانه کاربر را تعیین کند: %w",
  "could_not_stat_env_file": "نتوانست وضعیت فایل .env را بررسی کند: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "تنظیم top P",
  "stream_help": "پخش زنده",
  "set_presence_penalty": "تنظیم جریمه حضور",
-  "use_model_defaults_raw_help": "استفاده از پیش‌فرض‌های مدل بدون ارسال گزینه‌های گفتگو (مثل دما و غیره) و استفاده از نقش کاربر به جای نقش سیستم برای الگوها.",
+  "use_model_defaults_raw_help": "از مقادیر پیش‌فرض مدل بدون ارسال گزینه‌های چت (temperature، top_p و غیره) استفاده می‌کند. فقط بر ارائه‌دهندگان سازگار با OpenAI تأثیر می‌گذارد. مدل‌های Anthropic همواره برای رعایت نیازهای خاص هر مدل از انتخاب هوشمند پارامتر استفاده می‌کنند.",
  "set_frequency_penalty": "تنظیم جریمه فرکانس",
  "list_all_patterns": "فهرست تمام الگوها",
  "list_all_available_models": "فهرست تمام مدل‌های موجود",
--- a/internal/i18n/locales/fr.json
+++ b/internal/i18n/locales/fr.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "le fournisseur %s ne prend pas en charge la transcription audio",
  "transcription_model_required": "un modèle de transcription est requis (utilisez --transcribe-model)",
  "youtube_not_configured": "YouTube n'est pas configuré, veuillez exécuter la procédure de configuration",
+  "youtube_api_key_required": "Clé API YouTube requise pour les commentaires et métadonnées. Exécutez 'fabric --setup' pour configurer",
+  "youtube_ytdlp_not_found": "yt-dlp introuvable dans PATH. Veuillez installer yt-dlp pour utiliser la fonctionnalité de transcription YouTube",
+  "youtube_invalid_url": "URL YouTube invalide, impossible d'obtenir l'ID de vidéo ou de liste de lecture : '%s'",
+  "youtube_url_is_playlist_not_video": "L'URL est une liste de lecture, pas une vidéo",
+  "youtube_no_video_id_found": "aucun ID de vidéo trouvé dans l'URL",
+  "youtube_rate_limit_exceeded": "Limite de taux YouTube dépassée. Réessayez plus tard ou utilisez différents arguments yt-dlp comme '--sleep-requests 1' pour ralentir les requêtes.",
+  "youtube_auth_required_bot_detection": "YouTube nécessite une authentification (détection de bot). Utilisez --yt-dlp-args='--cookies-from-browser BROWSER' où BROWSER peut être chrome, firefox, brave, etc.",
+  "youtube_ytdlp_stderr_error": "Erreur lors de la lecture du stderr de yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "arguments yt-dlp invalides : %v",
+  "youtube_failed_create_temp_dir": "échec de création du répertoire temporaire : %v",
+  "youtube_no_transcript_content": "aucun contenu de transcription trouvé dans le fichier VTT",
+  "youtube_no_vtt_files_found": "aucun fichier VTT trouvé dans le répertoire",
+  "youtube_failed_walk_directory": "échec du parcours du répertoire : %v",
+  "youtube_error_getting_video_details": "erreur lors de l'obtention des détails de la vidéo : %v",
+  "youtube_invalid_duration_string": "chaîne de durée invalide : %s",
+  "youtube_error_getting_metadata": "erreur lors de l'obtention des métadonnées de la vidéo : %v",
+  "youtube_error_parsing_duration": "erreur lors de l'analyse de la durée de la vidéo : %v",
+  "youtube_error_getting_comments": "erreur lors de l'obtention des commentaires : %v",
+  "youtube_error_saving_csv": "erreur lors de l'enregistrement des vidéos en CSV : %v",
+  "youtube_no_video_found_with_id": "aucune vidéo trouvée avec l'ID : %s",
+  "youtube_invalid_timestamp_format": "format d'horodatage invalide : %s",
+  "youtube_empty_seconds_string": "chaîne de secondes vide",
+  "youtube_invalid_seconds_format": "format de secondes invalide %q : %w",
  "error_fetching_playlist_videos": "erreur lors de la récupération des vidéos de la liste de lecture : %w",
+  "openai_api_base_url_not_configured": "URL de base de l'API non configurée pour le fournisseur %s",
+  "openai_failed_to_create_models_url": "échec de création de l'URL des modèles : %w",
+  "openai_unexpected_status_code_with_body": "code d'état inattendu : %d du fournisseur %s, corps de réponse : %s",
+  "openai_unexpected_status_code_read_error_partial": "code d'état inattendu : %d du fournisseur %s (erreur de lecture : %v), réponse partielle : %s",
+  "openai_unexpected_status_code_read_error": "code d'état inattendu : %d du fournisseur %s (échec de lecture du corps de réponse : %v)",
+  "openai_unable_to_parse_models_response": "impossible d'analyser la réponse des modèles ; réponse brute : %s",
  "scraping_not_configured": "la fonctionnalité de scraping n'est pas configurée. Veuillez configurer Jina pour activer le scraping",
  "could_not_determine_home_dir": "impossible de déterminer le répertoire home de l'utilisateur : %w",
  "could_not_stat_env_file": "impossible de vérifier le fichier .env : %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Définir le top P",
  "stream_help": "Streaming",
  "set_presence_penalty": "Définir la pénalité de présence",
-  "use_model_defaults_raw_help": "Utiliser les valeurs par défaut du modèle sans envoyer d'options de chat (comme la température, etc.) et utiliser le rôle utilisateur au lieu du rôle système pour les motifs.",
+  "use_model_defaults_raw_help": "Utilise les valeurs par défaut du modèle sans envoyer d’options de discussion (temperature, top_p, etc.). N’affecte que les fournisseurs compatibles avec OpenAI. Les modèles Anthropic utilisent toujours une sélection intelligente des paramètres pour respecter les exigences propres à chaque modèle.",
  "set_frequency_penalty": "Définir la pénalité de fréquence",
  "list_all_patterns": "Lister tous les motifs",
  "list_all_available_models": "Lister tous les modèles disponibles",
--- a/internal/i18n/locales/it.json
+++ b/internal/i18n/locales/it.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "il fornitore %s non supporta la trascrizione audio",
  "transcription_model_required": "è richiesto un modello di trascrizione (usa --transcribe-model)",
  "youtube_not_configured": "YouTube non è configurato, per favore esegui la procedura di configurazione",
+  "youtube_api_key_required": "Chiave API YouTube richiesta per commenti e metadati. Esegui 'fabric --setup' per configurare",
+  "youtube_ytdlp_not_found": "yt-dlp non trovato in PATH. Per favore installa yt-dlp per usare la funzionalità di trascrizione YouTube",
+  "youtube_invalid_url": "URL YouTube non valido, impossibile ottenere l'ID del video o della playlist: '%s'",
+  "youtube_url_is_playlist_not_video": "L'URL è una playlist, non un video",
+  "youtube_no_video_id_found": "nessun ID video trovato nell'URL",
+  "youtube_rate_limit_exceeded": "Limite di richieste YouTube superato. Riprova più tardi o usa argomenti yt-dlp diversi come '--sleep-requests 1' per rallentare le richieste.",
+  "youtube_auth_required_bot_detection": "YouTube richiede autenticazione (rilevamento bot). Usa --yt-dlp-args='--cookies-from-browser BROWSER' dove BROWSER può essere chrome, firefox, brave, ecc.",
+  "youtube_ytdlp_stderr_error": "Errore durante la lettura dello stderr di yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "argomenti yt-dlp non validi: %v",
+  "youtube_failed_create_temp_dir": "impossibile creare la directory temporanea: %v",
+  "youtube_no_transcript_content": "nessun contenuto di trascrizione trovato nel file VTT",
+  "youtube_no_vtt_files_found": "nessun file VTT trovato nella directory",
+  "youtube_failed_walk_directory": "impossibile esplorare la directory: %v",
+  "youtube_error_getting_video_details": "errore nell'ottenere i dettagli del video: %v",
+  "youtube_invalid_duration_string": "stringa di durata non valida: %s",
+  "youtube_error_getting_metadata": "errore nell'ottenere i metadati del video: %v",
+  "youtube_error_parsing_duration": "errore nell'analizzare la durata del video: %v",
+  "youtube_error_getting_comments": "errore nell'ottenere i commenti: %v",
+  "youtube_error_saving_csv": "errore nel salvare i video in CSV: %v",
+  "youtube_no_video_found_with_id": "nessun video trovato con ID: %s",
+  "youtube_invalid_timestamp_format": "formato timestamp non valido: %s",
+  "youtube_empty_seconds_string": "stringa di secondi vuota",
+  "youtube_invalid_seconds_format": "formato secondi non valido %q: %w",
  "error_fetching_playlist_videos": "errore nel recupero dei video della playlist: %w",
+  "openai_api_base_url_not_configured": "URL base API non configurato per il provider %s",
+  "openai_failed_to_create_models_url": "impossibile creare URL modelli: %w",
+  "openai_unexpected_status_code_with_body": "codice di stato imprevisto: %d dal provider %s, corpo risposta: %s",
+  "openai_unexpected_status_code_read_error_partial": "codice di stato imprevisto: %d dal provider %s (errore lettura corpo: %v), risposta parziale: %s",
+  "openai_unexpected_status_code_read_error": "codice di stato imprevisto: %d dal provider %s (errore lettura corpo risposta: %v)",
+  "openai_unable_to_parse_models_response": "impossibile analizzare risposta modelli; risposta grezza: %s",
  "scraping_not_configured": "la funzionalità di scraping non è configurata. Per favore configura Jina per abilitare lo scraping",
  "could_not_determine_home_dir": "impossibile determinare la directory home dell'utente: %w",
  "could_not_stat_env_file": "impossibile verificare il file .env: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Imposta top P",
  "stream_help": "Streaming",
  "set_presence_penalty": "Imposta penalità di presenza",
-  "use_model_defaults_raw_help": "Usa i valori predefiniti del modello senza inviare opzioni di chat (come temperatura, ecc.) e usa il ruolo utente invece del ruolo sistema per i pattern.",
+  "use_model_defaults_raw_help": "Usa i valori predefiniti del modello senza inviare opzioni della chat (temperature, top_p, ecc.). Si applica solo ai provider compatibili con OpenAI. I modelli Anthropic utilizzano sempre una selezione intelligente dei parametri per rispettare i requisiti specifici del modello.",
  "set_frequency_penalty": "Imposta penalità di frequenza",
  "list_all_patterns": "Elenca tutti i pattern",
  "list_all_available_models": "Elenca tutti i modelli disponibili",
--- a/internal/i18n/locales/ja.json
+++ b/internal/i18n/locales/ja.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "ベンダー %s は音声転写をサポートしていません",
  "transcription_model_required": "転写モデルが必要です（--transcribe-model を使用）",
  "youtube_not_configured": "YouTubeが設定されていません。セットアップ手順を実行してください",
+  "youtube_api_key_required": "コメントとメタデータにはYouTube APIキーが必要です。設定するには 'fabric --setup' を実行してください",
+  "youtube_ytdlp_not_found": "PATHにyt-dlpが見つかりません。YouTubeトランスクリプト機能を使用するにはyt-dlpをインストールしてください",
+  "youtube_invalid_url": "無効なYouTube URL、動画またはプレイリストIDを取得できません: '%s'",
+  "youtube_url_is_playlist_not_video": "URLはプレイリストであり、動画ではありません",
+  "youtube_no_video_id_found": "URLに動画IDが見つかりません",
+  "youtube_rate_limit_exceeded": "YouTubeのレート制限を超えました。後でもう一度試すか、'--sleep-requests 1'のような異なるyt-dlp引数を使用してリクエストを遅くしてください。",
+  "youtube_auth_required_bot_detection": "YouTubeは認証を必要としています（ボット検出）。--yt-dlp-args='--cookies-from-browser BROWSER'を使用してください。BROWSERはchrome、firefox、braveなどです。",
+  "youtube_ytdlp_stderr_error": "yt-dlp stderrの読み取りエラー",
+  "youtube_invalid_ytdlp_arguments": "無効なyt-dlp引数: %v",
+  "youtube_failed_create_temp_dir": "一時ディレクトリの作成に失敗しました: %v",
+  "youtube_no_transcript_content": "VTTファイルにトランスクリプトコンテンツが見つかりません",
+  "youtube_no_vtt_files_found": "ディレクトリにVTTファイルが見つかりません",
+  "youtube_failed_walk_directory": "ディレクトリの走査に失敗しました: %v",
+  "youtube_error_getting_video_details": "動画の詳細取得エラー: %v",
+  "youtube_invalid_duration_string": "無効な長さ文字列: %s",
+  "youtube_error_getting_metadata": "動画のメタデータ取得エラー: %v",
+  "youtube_error_parsing_duration": "動画の長さ解析エラー: %v",
+  "youtube_error_getting_comments": "コメント取得エラー: %v",
+  "youtube_error_saving_csv": "動画のCSV保存エラー: %v",
+  "youtube_no_video_found_with_id": "IDの動画が見つかりません: %s",
+  "youtube_invalid_timestamp_format": "無効なタイムスタンプ形式: %s",
+  "youtube_empty_seconds_string": "空の秒文字列",
+  "youtube_invalid_seconds_format": "無効な秒形式 %q: %w",
  "error_fetching_playlist_videos": "プレイリスト動画の取得エラー: %w",
+  "openai_api_base_url_not_configured": "プロバイダー %s のAPIベースURLが設定されていません",
+  "openai_failed_to_create_models_url": "モデルURLの作成に失敗しました: %w",
+  "openai_unexpected_status_code_with_body": "予期しないステータスコード: プロバイダー %s から %d、レスポンス本文: %s",
+  "openai_unexpected_status_code_read_error_partial": "予期しないステータスコード: プロバイダー %s から %d (本文読み取りエラー: %v)、部分的なレスポンス: %s",
+  "openai_unexpected_status_code_read_error": "予期しないステータスコード: プロバイダー %s から %d (レスポンス本文の読み取りに失敗: %v)",
+  "openai_unable_to_parse_models_response": "モデルレスポンスの解析に失敗しました; 生のレスポンス: %s",
  "scraping_not_configured": "スクレイピング機能が設定されていません。スクレイピングを有効にするためにJinaを設定してください",
  "could_not_determine_home_dir": "ユーザーのホームディレクトリを特定できませんでした: %w",
  "could_not_stat_env_file": ".envファイルの状態を確認できませんでした: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Top Pを設定",
  "stream_help": "ストリーミング",
  "set_presence_penalty": "プレゼンスペナルティを設定",
-  "use_model_defaults_raw_help": "チャットオプション（温度など）を送信せずにモデルのデフォルトを使用し、パターンにシステムロールではなくユーザーロールを使用します。",
+  "use_model_defaults_raw_help": "チャットオプション（temperature、top_p など）を送信せずにモデルのデフォルトを使用します。OpenAI 互換プロバイダーにのみ適用されます。Anthropic モデルは常に、モデル固有の要件に準拠するためにスマートなパラメーター選択を使用します。",
  "set_frequency_penalty": "頻度ペナルティを設定",
  "list_all_patterns": "すべてのパターンを一覧表示",
  "list_all_available_models": "すべての利用可能なモデルを一覧表示",
--- a/internal/i18n/locales/pt-BR.json
+++ b/internal/i18n/locales/pt-BR.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "o fornecedor %s não suporta transcrição de áudio",
  "transcription_model_required": "modelo de transcrição é necessário (use --transcribe-model)",
  "youtube_not_configured": "YouTube não está configurado, por favor execute o procedimento de configuração",
+  "youtube_api_key_required": "Chave de API do YouTube necessária para comentários e metadados. Execute 'fabric --setup' para configurar",
+  "youtube_ytdlp_not_found": "yt-dlp não encontrado no PATH. Por favor instale o yt-dlp para usar a funcionalidade de transcrição do YouTube",
+  "youtube_invalid_url": "URL do YouTube inválida, não é possível obter o ID do vídeo ou da playlist: '%s'",
+  "youtube_url_is_playlist_not_video": "A URL é uma playlist, não um vídeo",
+  "youtube_no_video_id_found": "nenhum ID de vídeo encontrado na URL",
+  "youtube_rate_limit_exceeded": "Limite de taxa do YouTube excedido. Tente novamente mais tarde ou use argumentos diferentes do yt-dlp como '--sleep-requests 1' para desacelerar as requisições.",
+  "youtube_auth_required_bot_detection": "YouTube requer autenticação (detecção de bot). Use --yt-dlp-args='--cookies-from-browser BROWSER' onde BROWSER pode ser chrome, firefox, brave, etc.",
+  "youtube_ytdlp_stderr_error": "Erro ao ler stderr do yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "argumentos do yt-dlp inválidos: %v",
+  "youtube_failed_create_temp_dir": "falha ao criar diretório temporário: %v",
+  "youtube_no_transcript_content": "nenhum conteúdo de transcrição encontrado no arquivo VTT",
+  "youtube_no_vtt_files_found": "nenhum arquivo VTT encontrado no diretório",
+  "youtube_failed_walk_directory": "falha ao percorrer o diretório: %v",
+  "youtube_error_getting_video_details": "erro ao obter detalhes do vídeo: %v",
+  "youtube_invalid_duration_string": "string de duração inválida: %s",
+  "youtube_error_getting_metadata": "erro ao obter metadados do vídeo: %v",
+  "youtube_error_parsing_duration": "erro ao analisar a duração do vídeo: %v",
+  "youtube_error_getting_comments": "erro ao obter comentários: %v",
+  "youtube_error_saving_csv": "erro ao salvar vídeos em CSV: %v",
+  "youtube_no_video_found_with_id": "nenhum vídeo encontrado com o ID: %s",
+  "youtube_invalid_timestamp_format": "formato de timestamp inválido: %s",
+  "youtube_empty_seconds_string": "string de segundos vazia",
+  "youtube_invalid_seconds_format": "formato de segundos inválido %q: %w",
  "error_fetching_playlist_videos": "erro ao buscar vídeos da playlist: %w",
+  "openai_api_base_url_not_configured": "URL base da API não configurada para o provedor %s",
+  "openai_failed_to_create_models_url": "falha ao criar URL de modelos: %w",
+  "openai_unexpected_status_code_with_body": "código de status inesperado: %d do provedor %s, corpo da resposta: %s",
+  "openai_unexpected_status_code_read_error_partial": "código de status inesperado: %d do provedor %s (erro ao ler corpo: %v), resposta parcial: %s",
+  "openai_unexpected_status_code_read_error": "código de status inesperado: %d do provedor %s (falha ao ler corpo da resposta: %v)",
+  "openai_unable_to_parse_models_response": "não foi possível analisar a resposta de modelos; resposta bruta: %s",
  "scraping_not_configured": "funcionalidade de scraping não está configurada. Por favor configure o Jina para ativar o scraping",
  "could_not_determine_home_dir": "não foi possível determinar o diretório home do usuário: %w",
  "could_not_stat_env_file": "não foi possível verificar o arquivo .env: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Definir top P",
  "stream_help": "Streaming",
  "set_presence_penalty": "Definir penalidade de presença",
-  "use_model_defaults_raw_help": "Usar as configurações padrão do modelo sem enviar opções de chat (como temperatura, etc.) e usar o papel de usuário em vez do papel de sistema para padrões.",
+  "use_model_defaults_raw_help": "Usa os padrões do modelo sem enviar opções de chat (temperature, top_p etc.). Afeta apenas provedores compatíveis com o OpenAI. Os modelos da Anthropic sempre utilizam seleção inteligente de parâmetros para cumprir os requisitos específicos de cada modelo.",
  "set_frequency_penalty": "Definir penalidade de frequência",
  "list_all_patterns": "Listar todos os padrões/patterns",
  "list_all_available_models": "Listar todos os modelos disponíveis",
@@ -133,4 +162,4 @@
  "no_description_available": "Nenhuma descrição disponível",
  "i18n_download_failed": "Falha ao baixar tradução para o idioma '%s': %v",
  "i18n_load_failed": "Falha ao carregar arquivo de tradução: %v"
-}
+}
--- a/internal/i18n/locales/pt-PT.json
+++ b/internal/i18n/locales/pt-PT.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "o fornecedor %s não suporta transcrição de áudio",
  "transcription_model_required": "modelo de transcrição é necessário (use --transcribe-model)",
  "youtube_not_configured": "YouTube não está configurado, por favor execute o procedimento de configuração",
+  "youtube_api_key_required": "Chave de API do YouTube necessária para comentários e metadados. Execute 'fabric --setup' para configurar",
+  "youtube_ytdlp_not_found": "yt-dlp não encontrado no PATH. Por favor instale o yt-dlp para usar a funcionalidade de transcrição do YouTube",
+  "youtube_invalid_url": "URL do YouTube inválido, não é possível obter o ID do vídeo ou da lista de reprodução: '%s'",
+  "youtube_url_is_playlist_not_video": "O URL é uma lista de reprodução, não um vídeo",
+  "youtube_no_video_id_found": "nenhum ID de vídeo encontrado no URL",
+  "youtube_rate_limit_exceeded": "Limite de taxa do YouTube excedido. Tente novamente mais tarde ou utilize argumentos diferentes do yt-dlp como '--sleep-requests 1' para desacelerar os pedidos.",
+  "youtube_auth_required_bot_detection": "YouTube requer autenticação (deteção de bot). Use --yt-dlp-args='--cookies-from-browser BROWSER' onde BROWSER pode ser chrome, firefox, brave, etc.",
+  "youtube_ytdlp_stderr_error": "Erro ao ler stderr do yt-dlp",
+  "youtube_invalid_ytdlp_arguments": "argumentos do yt-dlp inválidos: %v",
+  "youtube_failed_create_temp_dir": "falha ao criar diretório temporário: %v",
+  "youtube_no_transcript_content": "nenhum conteúdo de transcrição encontrado no ficheiro VTT",
+  "youtube_no_vtt_files_found": "nenhum ficheiro VTT encontrado no diretório",
+  "youtube_failed_walk_directory": "falha ao percorrer o diretório: %v",
+  "youtube_error_getting_video_details": "erro ao obter detalhes do vídeo: %v",
+  "youtube_invalid_duration_string": "cadeia de duração inválida: %s",
+  "youtube_error_getting_metadata": "erro ao obter metadados do vídeo: %v",
+  "youtube_error_parsing_duration": "erro ao analisar a duração do vídeo: %v",
+  "youtube_error_getting_comments": "erro ao obter comentários: %v",
+  "youtube_error_saving_csv": "erro ao guardar vídeos em CSV: %v",
+  "youtube_no_video_found_with_id": "nenhum vídeo encontrado com o ID: %s",
+  "youtube_invalid_timestamp_format": "formato de timestamp inválido: %s",
+  "youtube_empty_seconds_string": "cadeia de segundos vazia",
+  "youtube_invalid_seconds_format": "formato de segundos inválido %q: %w",
  "error_fetching_playlist_videos": "erro ao obter vídeos da playlist: %w",
+  "openai_api_base_url_not_configured": "URL base da API não configurado para o fornecedor %s",
+  "openai_failed_to_create_models_url": "falha ao criar URL de modelos: %w",
+  "openai_unexpected_status_code_with_body": "código de estado inesperado: %d do fornecedor %s, corpo da resposta: %s",
+  "openai_unexpected_status_code_read_error_partial": "código de estado inesperado: %d do fornecedor %s (erro ao ler corpo: %v), resposta parcial: %s",
+  "openai_unexpected_status_code_read_error": "código de estado inesperado: %d do fornecedor %s (falha ao ler corpo da resposta: %v)",
+  "openai_unable_to_parse_models_response": "não foi possível analisar a resposta de modelos; resposta bruta: %s",
  "scraping_not_configured": "funcionalidade de scraping não está configurada. Por favor configure o Jina para ativar o scraping",
  "could_not_determine_home_dir": "não foi possível determinar o diretório home do utilizador: %w",
  "could_not_stat_env_file": "não foi possível verificar o ficheiro .env: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "Definir top P",
  "stream_help": "Streaming",
  "set_presence_penalty": "Definir penalidade de presença",
-  "use_model_defaults_raw_help": "Usar as predefinições do modelo sem enviar opções de chat (como temperatura, etc.) e usar o papel de utilizador em vez do papel de sistema para padrões.",
+  "use_model_defaults_raw_help": "Utiliza os valores predefinidos do modelo sem enviar opções de chat (temperature, top_p, etc.). Só afeta fornecedores compatíveis com o OpenAI. Os modelos Anthropic usam sempre uma seleção inteligente de parâmetros para cumprir os requisitos específicos do modelo.",
  "set_frequency_penalty": "Definir penalidade de frequência",
  "list_all_patterns": "Listar todos os padrões",
  "list_all_available_models": "Listar todos os modelos disponíveis",
--- a/internal/i18n/locales/zh.json
+++ b/internal/i18n/locales/zh.json
@@ -4,7 +4,36 @@
  "vendor_no_transcription_support": "供应商 %s 不支持音频转录",
  "transcription_model_required": "需要转录模型（使用 --transcribe-model）",
  "youtube_not_configured": "YouTube 未配置，请运行设置程序",
+  "youtube_api_key_required": "评论和元数据需要 YouTube API 密钥。运行 'fabric --setup' 进行配置",
+  "youtube_ytdlp_not_found": "在 PATH 中未找到 yt-dlp。请安装 yt-dlp 以使用 YouTube 转录功能",
+  "youtube_invalid_url": "无效的 YouTube URL，无法获取视频或播放列表 ID：'%s'",
+  "youtube_url_is_playlist_not_video": "URL 是播放列表，而不是视频",
+  "youtube_no_video_id_found": "在 URL 中未找到视频 ID",
+  "youtube_rate_limit_exceeded": "超过 YouTube 速率限制。请稍后重试，或使用不同的 yt-dlp 参数（如 '--sleep-requests 1'）来减慢请求速度。",
+  "youtube_auth_required_bot_detection": "YouTube 需要身份验证（机器人检测）。使用 --yt-dlp-args='--cookies-from-browser BROWSER'，其中 BROWSER 可以是 chrome、firefox、brave 等。",
+  "youtube_ytdlp_stderr_error": "读取 yt-dlp stderr 时出错",
+  "youtube_invalid_ytdlp_arguments": "无效的 yt-dlp 参数：%v",
+  "youtube_failed_create_temp_dir": "创建临时目录失败：%v",
+  "youtube_no_transcript_content": "在 VTT 文件中未找到转录内容",
+  "youtube_no_vtt_files_found": "在目录中未找到 VTT 文件",
+  "youtube_failed_walk_directory": "遍历目录失败：%v",
+  "youtube_error_getting_video_details": "获取视频详情时出错：%v",
+  "youtube_invalid_duration_string": "无效的时长字符串：%s",
+  "youtube_error_getting_metadata": "获取视频元数据时出错：%v",
+  "youtube_error_parsing_duration": "解析视频时长时出错：%v",
+  "youtube_error_getting_comments": "获取评论时出错：%v",
+  "youtube_error_saving_csv": "将视频保存为 CSV 时出错：%v",
+  "youtube_no_video_found_with_id": "未找到 ID 为 %s 的视频",
+  "youtube_invalid_timestamp_format": "无效的时间戳格式：%s",
+  "youtube_empty_seconds_string": "秒数字符串为空",
+  "youtube_invalid_seconds_format": "无效的秒数格式 %q：%w",
  "error_fetching_playlist_videos": "获取播放列表视频时出错: %w",
+  "openai_api_base_url_not_configured": "未为提供商 %s 配置 API 基础 URL",
+  "openai_failed_to_create_models_url": "创建模型 URL 失败：%w",
+  "openai_unexpected_status_code_with_body": "意外的状态码：来自提供商 %s 的 %d，响应主体：%s",
+  "openai_unexpected_status_code_read_error_partial": "意外的状态码：来自提供商 %s 的 %d（读取主体错误：%v），部分响应：%s",
+  "openai_unexpected_status_code_read_error": "意外的状态码：来自提供商 %s 的 %d（读取响应主体失败：%v)",
+  "openai_unable_to_parse_models_response": "无法解析模型响应；原始响应：%s",
  "scraping_not_configured": "抓取功能未配置。请设置 Jina 以启用抓取功能",
  "could_not_determine_home_dir": "无法确定用户主目录: %w",
  "could_not_stat_env_file": "无法获取 .env 文件状态: %w",
@@ -53,7 +82,7 @@
  "set_top_p": "设置 top P",
  "stream_help": "流式传输",
  "set_presence_penalty": "设置存在惩罚",
-  "use_model_defaults_raw_help": "使用模型默认设置，不发送聊天选项（如温度等），对于模式使用用户角色而非系统角色。",
+  "use_model_defaults_raw_help": "在不发送聊天选项（temperature、top_p 等）的情况下使用模型默认值。仅影响兼容 OpenAI 的提供商。Anthropic 模型始终使用智能参数选择以满足特定模型的要求。",
  "set_frequency_penalty": "设置频率惩罚",
  "list_all_patterns": "列出所有模式",
  "list_all_available_models": "列出所有可用模型",
--- a/internal/log/log.go
+++ b/internal/log/log.go
@@ -51,7 +51,7 @@ func LevelFromInt(i int) Level {
 }

 // Debug writes a debug message if the global level permits.
-func Debug(l Level, format string, a ...interface{}) {
+func Debug(l Level, format string, a ...any) {
 	mu.RLock()
 	current := level
 	w := output
@@ -63,7 +63,7 @@ func Debug(l Level, format string, a ...interface{}) {

 // Log writes a message unconditionally to stderr.
 // This is for important messages that should always be shown regardless of debug level.
-func Log(format string, a ...interface{}) {
+func Log(format string, a ...any) {
 	mu.RLock()
 	w := output
 	mu.RUnlock()
--- a/internal/plugins/ai/anthropic/anthropic.go
+++ b/internal/plugins/ai/anthropic/anthropic.go
@@ -4,6 +4,7 @@ import (
 	"context"
 	"fmt"
 	"net/http"
+	"os"
 	"strconv"
 	"strings"

@@ -44,17 +45,22 @@ func NewClient() (ret *Client) {
 	ret.models = []string{
 		string(anthropic.ModelClaude3_7SonnetLatest), string(anthropic.ModelClaude3_7Sonnet20250219),
 		string(anthropic.ModelClaude3_5HaikuLatest), string(anthropic.ModelClaude3_5Haiku20241022),
-		string(anthropic.ModelClaude3_5SonnetLatest), string(anthropic.ModelClaude3_5Sonnet20241022),
-		string(anthropic.ModelClaude_3_5_Sonnet_20240620), string(anthropic.ModelClaude3OpusLatest),
-		string(anthropic.ModelClaude_3_Opus_20240229), string(anthropic.ModelClaude_3_Haiku_20240307),
+		string(anthropic.ModelClaude3OpusLatest), string(anthropic.ModelClaude_3_Opus_20240229),
+		string(anthropic.ModelClaude_3_Haiku_20240307),
 		string(anthropic.ModelClaudeOpus4_20250514), string(anthropic.ModelClaudeSonnet4_20250514),
 		string(anthropic.ModelClaudeOpus4_1_20250805),
 		string(anthropic.ModelClaudeSonnet4_5),
+		string(anthropic.ModelClaudeSonnet4_5_20250929),
+		string(anthropic.ModelClaudeOpus4_5_20251101),
+		string(anthropic.ModelClaudeOpus4_5),
+		string(anthropic.ModelClaudeHaiku4_5),
+		string(anthropic.ModelClaudeHaiku4_5_20251001),
 	}

 	ret.modelBetas = map[string][]string{
-		string(anthropic.ModelClaudeSonnet4_20250514): {"context-1m-2025-08-07"},
-		string(anthropic.ModelClaudeSonnet4_5):        {"context-1m-2025-08-07"},
+		string(anthropic.ModelClaudeSonnet4_20250514):   {"context-1m-2025-08-07"},
+		string(anthropic.ModelClaudeSonnet4_5):          {"context-1m-2025-08-07"},
+		string(anthropic.ModelClaudeSonnet4_5_20250929): {"context-1m-2025-08-07"},
 	}

 	return
@@ -211,7 +217,7 @@ func (an *Client) SendStream(
 	}

 	if stream.Err() != nil {
-		fmt.Printf("Messages stream error: %v\n", stream.Err())
+		fmt.Fprintf(os.Stderr, "Messages stream error: %v\n", stream.Err())
 	}
 	close(channel)
 	return
@@ -355,7 +361,7 @@ func (an *Client) toMessages(msgs []*chat.ChatCompletionMessage) (ret []anthropi
 	lastRoleWasUser := false

 	for _, msg := range msgs {
-		if msg.Content == "" {
+		if strings.TrimSpace(msg.Content) == "" {
 			continue // Skip empty messages
 		}

--- a/internal/plugins/ai/anthropic/oauth_test.go
+++ b/internal/plugins/ai/anthropic/oauth_test.go
@@ -52,7 +52,7 @@ func createExpiredToken(accessToken, refreshToken string) *util.OAuthToken {
 }

 // mockTokenServer creates a mock OAuth token server for testing
-func mockTokenServer(_ *testing.T, responses map[string]interface{}) *httptest.Server {
+func mockTokenServer(_ *testing.T, responses map[string]any) *httptest.Server {
 	return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		if r.URL.Path != "/v1/oauth/token" {
 			http.NotFound(w, r)
@@ -80,7 +80,7 @@ func mockTokenServer(_ *testing.T, responses map[string]interface{}) *httptest.S

 		w.Header().Set("Content-Type", "application/json")

-		if errorResp, ok := response.(map[string]interface{}); ok && errorResp["error"] != nil {
+		if errorResp, ok := response.(map[string]any); ok && errorResp["error"] != nil {
 			w.WriteHeader(http.StatusBadRequest)
 		}

@@ -114,8 +114,8 @@ func TestGeneratePKCE(t *testing.T) {

 func TestExchangeToken_Success(t *testing.T) {
 	// Create mock server
-	server := mockTokenServer(t, map[string]interface{}{
-		"authorization_code": map[string]interface{}{
+	server := mockTokenServer(t, map[string]any{
+		"authorization_code": map[string]any{
 			"access_token":  "test_access_token",
 			"refresh_token": "test_refresh_token",
 			"expires_in":    3600,
@@ -161,8 +161,8 @@ func TestRefreshToken_Success(t *testing.T) {
 	os.WriteFile(tokenPath, data, 0600)

 	// Create mock server for refresh
-	server := mockTokenServer(t, map[string]interface{}{
-		"refresh_token": map[string]interface{}{
+	server := mockTokenServer(t, map[string]any{
+		"refresh_token": map[string]any{
 			"access_token":  "new_access_token",
 			"refresh_token": "new_refresh_token",
 			"expires_in":    3600,
@@ -416,7 +416,7 @@ func TestGetValidTokenWithValidToken(t *testing.T) {

 // Benchmark tests
 func BenchmarkGeneratePKCE(b *testing.B) {
-	for i := 0; i < b.N; i++ {
+	for b.Loop() {
 		_, _, err := generatePKCE()
 		if err != nil {
 			b.Fatal(err)
@@ -427,8 +427,7 @@ func BenchmarkGeneratePKCE(b *testing.B) {
 func BenchmarkTokenIsExpired(b *testing.B) {
 	token := createTestToken("access", "refresh", 3600)

-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
+	for b.Loop() {
 		token.IsExpired(5)
 	}
 }
--- a/internal/plugins/ai/gemini/gemini.go
+++ b/internal/plugins/ai/gemini/gemini.go
@@ -131,6 +131,8 @@ func (o *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, o

 func (o *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions, channel chan string) (err error) {
 	ctx := context.Background()
+	defer close(channel)
+
 	var client *genai.Client
 	if client, err = genai.NewClient(ctx, &genai.ClientConfig{
 		APIKey:  o.ApiKey.Value,
@@ -153,8 +155,7 @@ func (o *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.Cha
 	for response, err := range stream {
 		if err != nil {
 			channel <- fmt.Sprintf("Error: %v\n", err)
-			close(channel)
-			break
+			return err
 		}

 		text := o.extractTextFromResponse(response)
@@ -162,7 +163,6 @@ func (o *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.Cha
 			channel <- text
 		}
 	}
-	close(channel)

 	return
 }
@@ -456,7 +456,7 @@ func (o *Client) convertMessages(msgs []*chat.ChatCompletionMessage) []*genai.Co
 			content.Role = "user"
 		}

-		if msg.Content != "" {
+		if strings.TrimSpace(msg.Content) != "" {
 			content.Parts = append(content.Parts, &genai.Part{Text: msg.Content})
 		}

--- a/internal/plugins/ai/gemini/voices.go
+++ b/internal/plugins/ai/gemini/voices.go
@@ -3,6 +3,7 @@ package gemini
 import (
 	"fmt"
 	"sort"
+	"strings"
 )

 // GeminiVoice represents a Gemini TTS voice with its characteristics
@@ -126,16 +127,17 @@ func ListGeminiVoices(shellCompleteMode bool) string {
 	if shellCompleteMode {
 		// For shell completion, just return voice names
 		names := GetGeminiVoiceNames()
-		result := ""
+		var result strings.Builder
 		for _, name := range names {
-			result += name + "\n"
+			result.WriteString(name + "\n")
 		}
-		return result
+		return result.String()
 	}

 	// For human-readable output
 	voices := GetGeminiVoices()
-	result := "Available Gemini Text-to-Speech voices:\n\n"
+	var result strings.Builder
+	result.WriteString("Available Gemini Text-to-Speech voices:\n\n")

 	// Group by characteristics for better readability
 	groups := map[string][]GeminiVoice{
@@ -186,22 +188,22 @@ func ListGeminiVoices(shellCompleteMode bool) string {
 	// Output grouped voices
 	for groupName, groupVoices := range groups {
 		if len(groupVoices) > 0 {
-			result += fmt.Sprintf("%s:\n", groupName)
+			result.WriteString(fmt.Sprintf("%s:\n", groupName))
 			for _, voice := range groupVoices {
 				defaultStr := ""
 				if voice.Name == "Kore" {
 					defaultStr = " (default)"
 				}
-				result += fmt.Sprintf("  %-15s - %s%s\n", voice.Name, voice.Description, defaultStr)
+				result.WriteString(fmt.Sprintf("  %-15s - %s%s\n", voice.Name, voice.Description, defaultStr))
 			}
-			result += "\n"
+			result.WriteString("\n")
 		}
 	}

-	result += "Use --voice <voice_name> to select a specific voice.\n"
-	result += "Example: fabric --voice Charon -m gemini-2.5-flash-preview-tts -o output.wav \"Hello world\"\n"
+	result.WriteString("Use --voice <voice_name> to select a specific voice.\n")
+	result.WriteString("Example: fabric --voice Charon -m gemini-2.5-flash-preview-tts -o output.wav \"Hello world\"\n")

-	return result
+	return result.String()
 }

 // NOTE: This implementation maintains a curated list based on official Google documentation.
--- a/internal/plugins/ai/lmstudio/lmstudio.go
+++ b/internal/plugins/ai/lmstudio/lmstudio.go
@@ -90,7 +90,7 @@ func (c *Client) ListModels() ([]string, error) {
 func (c *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions, channel chan string) (err error) {
 	url := fmt.Sprintf("%s/chat/completions", c.ApiUrl.Value)

-	payload := map[string]interface{}{
+	payload := map[string]any{
 		"messages": msgs,
 		"model":    opts.Model,
 		"stream":   true, // Enable streaming
@@ -140,27 +140,27 @@ func (c *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.Cha
 			continue
 		}

-		if bytes.HasPrefix(line, []byte("data: ")) {
-			line = bytes.TrimPrefix(line, []byte("data: "))
+		if after, ok := bytes.CutPrefix(line, []byte("data: ")); ok {
+			line = after
 		}

 		if string(line) == "[DONE]" {
 			break
 		}

-		var result map[string]interface{}
+		var result map[string]any
 		if err = json.Unmarshal(line, &result); err != nil {
 			continue
 		}

-		var choices []interface{}
+		var choices []any
 		var ok bool
-		if choices, ok = result["choices"].([]interface{}); !ok || len(choices) == 0 {
+		if choices, ok = result["choices"].([]any); !ok || len(choices) == 0 {
 			continue
 		}

-		var delta map[string]interface{}
-		if delta, ok = choices[0].(map[string]interface{})["delta"].(map[string]interface{}); !ok {
+		var delta map[string]any
+		if delta, ok = choices[0].(map[string]any)["delta"].(map[string]any); !ok {
 			continue
 		}

@@ -176,7 +176,7 @@ func (c *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.Cha
 func (c *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions) (content string, err error) {
 	url := fmt.Sprintf("%s/chat/completions", c.ApiUrl.Value)

-	payload := map[string]interface{}{
+	payload := map[string]any{
 		"messages": msgs,
 		"model":    opts.Model,
 		// Add other options from opts if supported by LM Studio
@@ -208,21 +208,21 @@ func (c *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, o
 		return
 	}

-	var result map[string]interface{}
+	var result map[string]any
 	if err = json.NewDecoder(resp.Body).Decode(&result); err != nil {
 		err = fmt.Errorf("failed to decode response: %w", err)
 		return
 	}

-	var choices []interface{}
+	var choices []any
 	var ok bool
-	if choices, ok = result["choices"].([]interface{}); !ok || len(choices) == 0 {
+	if choices, ok = result["choices"].([]any); !ok || len(choices) == 0 {
 		err = fmt.Errorf("invalid response format: missing or empty choices")
 		return
 	}

-	var message map[string]interface{}
-	if message, ok = choices[0].(map[string]interface{})["message"].(map[string]interface{}); !ok {
+	var message map[string]any
+	if message, ok = choices[0].(map[string]any)["message"].(map[string]any); !ok {
 		err = fmt.Errorf("invalid response format: missing message in first choice")
 		return
 	}
@@ -238,7 +238,7 @@ func (c *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, o
 func (c *Client) Complete(ctx context.Context, prompt string, opts *domain.ChatOptions) (text string, err error) {
 	url := fmt.Sprintf("%s/completions", c.ApiUrl.Value)

-	payload := map[string]interface{}{
+	payload := map[string]any{
 		"prompt": prompt,
 		"model":  opts.Model,
 		// Add other options from opts if supported by LM Studio
@@ -270,20 +270,20 @@ func (c *Client) Complete(ctx context.Context, prompt string, opts *domain.ChatO
 		return
 	}

-	var result map[string]interface{}
+	var result map[string]any
 	if err = json.NewDecoder(resp.Body).Decode(&result); err != nil {
 		err = fmt.Errorf("failed to decode response: %w", err)
 		return
 	}

-	var choices []interface{}
+	var choices []any
 	var ok bool
-	if choices, ok = result["choices"].([]interface{}); !ok || len(choices) == 0 {
+	if choices, ok = result["choices"].([]any); !ok || len(choices) == 0 {
 		err = fmt.Errorf("invalid response format: missing or empty choices")
 		return
 	}

-	if text, ok = choices[0].(map[string]interface{})["text"].(string); !ok {
+	if text, ok = choices[0].(map[string]any)["text"].(string); !ok {
 		err = fmt.Errorf("invalid response format: missing or non-string text in first choice")
 		return
 	}
@@ -294,7 +294,7 @@ func (c *Client) Complete(ctx context.Context, prompt string, opts *domain.ChatO
 func (c *Client) GetEmbeddings(ctx context.Context, input string, opts *domain.ChatOptions) (embeddings []float64, err error) {
 	url := fmt.Sprintf("%s/embeddings", c.ApiUrl.Value)

-	payload := map[string]interface{}{
+	payload := map[string]any{
 		"input": input,
 		"model": opts.Model,
 		// Add other options from opts if supported by LM Studio
--- a/internal/plugins/ai/models.go
+++ b/internal/plugins/ai/models.go
@@ -17,6 +17,35 @@ type VendorsModels struct {
 	*util.GroupsItemsSelectorString
 }

+// FilterByVendor returns a new VendorsModels containing only the specified vendor's models.
+// Vendor matching is case-insensitive (e.g., "OpenAI", "openai", and "OPENAI" all match).
+// If the vendor is not found, an empty VendorsModels is returned.
+func (o *VendorsModels) FilterByVendor(vendor string) *VendorsModels {
+	filtered := NewVendorsModels()
+	for _, groupItems := range o.GroupsItems {
+		if strings.EqualFold(groupItems.Group, vendor) {
+			filtered.AddGroupItems(groupItems.Group, groupItems.Items...)
+			break
+		}
+	}
+	return filtered
+}
+
+// FindModelNameCaseInsensitive returns the actual model name from available models,
+// matching case-insensitively. Returns empty string if not found.
+// For example, if the available models contain "gpt-4o" and user queries "GPT-4O",
+// this returns "gpt-4o" (the actual model name that should be sent to the API).
+func (o *VendorsModels) FindModelNameCaseInsensitive(modelQuery string) string {
+	for _, groupItems := range o.GroupsItems {
+		for _, item := range groupItems.Items {
+			if strings.EqualFold(item, modelQuery) {
+				return item
+			}
+		}
+	}
+	return ""
+}
+
 // PrintWithVendor prints models including their vendor on each line.
 // When shellCompleteList is true, output is suitable for shell completion.
 // Default vendor and model are highlighted with an asterisk.
--- a/internal/plugins/ai/models_test.go
+++ b/internal/plugins/ai/models_test.go
@@ -19,19 +19,19 @@ func TestNewVendorsModels(t *testing.T) {

 func TestFindVendorsByModelFirst(t *testing.T) {
 	vendors := NewVendorsModels()
-	vendors.AddGroupItems("vendor1", []string{"model1", "model2"}...)
+	vendors.AddGroupItems("Vendor1", []string{"Model1", "model2"}...)
 	vendor := vendors.FindGroupsByItemFirst("model1")
-	if vendor != "vendor1" {
-		t.Fatalf("FindVendorsByModelFirst() = %v, want %v", vendor, "vendor1")
+	if vendor != "Vendor1" {
+		t.Fatalf("FindVendorsByModelFirst() = %v, want %v", vendor, "Vendor1")
 	}
 }

 func TestFindVendorsByModel(t *testing.T) {
 	vendors := NewVendorsModels()
-	vendors.AddGroupItems("vendor1", []string{"model1", "model2"}...)
-	foundVendors := vendors.FindGroupsByItem("model1")
-	if len(foundVendors) != 1 || foundVendors[0] != "vendor1" {
-		t.Fatalf("FindVendorsByModel() = %v, want %v", foundVendors, []string{"vendor1"})
+	vendors.AddGroupItems("Vendor1", []string{"Model1", "model2"}...)
+	foundVendors := vendors.FindGroupsByItem("MODEL1")
+	if len(foundVendors) != 1 || foundVendors[0] != "Vendor1" {
+		t.Fatalf("FindVendorsByModel() = %v, want %v", foundVendors, []string{"Vendor1"})
 	}
 }

@@ -54,3 +54,51 @@ func TestPrintWithVendorMarksDefault(t *testing.T) {
 		t.Fatalf("default model not marked: %s", out)
 	}
 }
+
+func TestFilterByVendorCaseInsensitive(t *testing.T) {
+	vendors := NewVendorsModels()
+	vendors.AddGroupItems("vendor1", []string{"model1"}...)
+	vendors.AddGroupItems("vendor2", []string{"model2"}...)
+
+	filtered := vendors.FilterByVendor("VENDOR2")
+
+	if len(filtered.GroupsItems) != 1 {
+		t.Fatalf("expected 1 vendor group, got %d", len(filtered.GroupsItems))
+	}
+
+	if filtered.GroupsItems[0].Group != "vendor2" {
+		t.Fatalf("expected vendor2, got %s", filtered.GroupsItems[0].Group)
+	}
+
+	if len(filtered.GroupsItems[0].Items) != 1 || filtered.GroupsItems[0].Items[0] != "model2" {
+		t.Fatalf("unexpected models for vendor2: %v", filtered.GroupsItems[0].Items)
+	}
+}
+
+func TestFindModelNameCaseInsensitive(t *testing.T) {
+	vendors := NewVendorsModels()
+	vendors.AddGroupItems("OpenAI", []string{"gpt-4o", "gpt-5"}...)
+	vendors.AddGroupItems("Anthropic", []string{"claude-3-opus"}...)
+
+	tests := []struct {
+		name          string
+		query         string
+		expectedModel string
+	}{
+		{"exact match lowercase", "gpt-4o", "gpt-4o"},
+		{"uppercase query", "GPT-4O", "gpt-4o"},
+		{"mixed case query", "GpT-5", "gpt-5"},
+		{"exact match with hyphens", "claude-3-opus", "claude-3-opus"},
+		{"uppercase with hyphens", "CLAUDE-3-OPUS", "claude-3-opus"},
+		{"non-existent model", "gpt-999", ""},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			result := vendors.FindModelNameCaseInsensitive(tt.query)
+			if result != tt.expectedModel {
+				t.Errorf("FindModelNameCaseInsensitive(%q) = %q, want %q", tt.query, result, tt.expectedModel)
+			}
+		})
+	}
+}
--- a/internal/plugins/ai/ollama/ollama.go
+++ b/internal/plugins/ai/ollama/ollama.go
@@ -2,7 +2,9 @@ package ollama

 import (
 	"context"
+	"encoding/base64"
 	"fmt"
+	"io"
 	"net/http"
 	"net/url"
 	"os"
@@ -10,11 +12,10 @@ import (
 	"time"

 	"github.com/danielmiessler/fabric/internal/chat"
-	ollamaapi "github.com/ollama/ollama/api"
-	"github.com/samber/lo"
-
 	"github.com/danielmiessler/fabric/internal/domain"
+	debuglog "github.com/danielmiessler/fabric/internal/log"
 	"github.com/danielmiessler/fabric/internal/plugins"
+	ollamaapi "github.com/ollama/ollama/api"
 )

 const defaultBaseUrl = "http://localhost:11434"
@@ -48,6 +49,7 @@ type Client struct {
 	apiUrl         *url.URL
 	client         *ollamaapi.Client
 	ApiHttpTimeout *plugins.SetupQuestion
+	httpClient     *http.Client
 }

 type transport_sec struct {
@@ -84,7 +86,8 @@ func (o *Client) configure() (err error) {
 		}
 	}

-	o.client = ollamaapi.NewClient(o.apiUrl, &http.Client{Timeout: timeout, Transport: &transport_sec{underlyingTransport: http.DefaultTransport, ApiKey: o.ApiKey}})
+	o.httpClient = &http.Client{Timeout: timeout, Transport: &transport_sec{underlyingTransport: http.DefaultTransport, ApiKey: o.ApiKey}}
+	o.client = ollamaapi.NewClient(o.apiUrl, o.httpClient)

 	return
 }
@@ -104,15 +107,18 @@ func (o *Client) ListModels() (ret []string, err error) {
 }

 func (o *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions, channel chan string) (err error) {
-	req := o.createChatRequest(msgs, opts)
+	ctx := context.Background()
+
+	var req ollamaapi.ChatRequest
+	if req, err = o.createChatRequest(ctx, msgs, opts); err != nil {
+		return
+	}

 	respFunc := func(resp ollamaapi.ChatResponse) (streamErr error) {
 		channel <- resp.Message.Content
 		return
 	}

-	ctx := context.Background()
-
 	if err = o.client.Chat(ctx, &req, respFunc); err != nil {
 		return
 	}
@@ -124,7 +130,10 @@ func (o *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.Cha
 func (o *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions) (ret string, err error) {
 	bf := false

-	req := o.createChatRequest(msgs, opts)
+	var req ollamaapi.ChatRequest
+	if req, err = o.createChatRequest(ctx, msgs, opts); err != nil {
+		return
+	}
 	req.Stream = &bf

 	respFunc := func(resp ollamaapi.ChatResponse) (streamErr error) {
@@ -133,17 +142,20 @@ func (o *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, o
 	}

 	if err = o.client.Chat(ctx, &req, respFunc); err != nil {
-		fmt.Printf("FRED --> %s\n", err)
+		debuglog.Debug(debuglog.Basic, "Ollama chat request failed: %v\n", err)
 	}
 	return
 }

-func (o *Client) createChatRequest(msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions) (ret ollamaapi.ChatRequest) {
-	messages := lo.Map(msgs, func(message *chat.ChatCompletionMessage, _ int) (ret ollamaapi.Message) {
-		return ollamaapi.Message{Role: message.Role, Content: message.Content}
-	})
+func (o *Client) createChatRequest(ctx context.Context, msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions) (ret ollamaapi.ChatRequest, err error) {
+	messages := make([]ollamaapi.Message, len(msgs))
+	for i, message := range msgs {
+		if messages[i], err = o.convertMessage(ctx, message); err != nil {
+			return
+		}
+	}

-	options := map[string]interface{}{
+	options := map[string]any{
 		"temperature":       opts.Temperature,
 		"presence_penalty":  opts.PresencePenalty,
 		"frequency_penalty": opts.FrequencyPenalty,
@@ -162,14 +174,85 @@ func (o *Client) createChatRequest(msgs []*chat.ChatCompletionMessage, opts *dom
 	return
 }

+func (o *Client) convertMessage(ctx context.Context, message *chat.ChatCompletionMessage) (ret ollamaapi.Message, err error) {
+	ret = ollamaapi.Message{Role: message.Role, Content: message.Content}
+
+	if len(message.MultiContent) == 0 {
+		return
+	}
+
+	// Pre-allocate with capacity hint
+	textParts := make([]string, 0, len(message.MultiContent))
+	if strings.TrimSpace(ret.Content) != "" {
+		textParts = append(textParts, strings.TrimSpace(ret.Content))
+	}
+
+	for _, part := range message.MultiContent {
+		switch part.Type {
+		case chat.ChatMessagePartTypeText:
+			if trimmed := strings.TrimSpace(part.Text); trimmed != "" {
+				textParts = append(textParts, trimmed)
+			}
+		case chat.ChatMessagePartTypeImageURL:
+			// Nil guard
+			if part.ImageURL == nil || part.ImageURL.URL == "" {
+				continue
+			}
+			var img []byte
+			if img, err = o.loadImageBytes(ctx, part.ImageURL.URL); err != nil {
+				return
+			}
+			ret.Images = append(ret.Images, ollamaapi.ImageData(img))
+		}
+	}
+
+	ret.Content = strings.Join(textParts, "\n")
+	return
+}
+
+func (o *Client) loadImageBytes(ctx context.Context, imageURL string) (ret []byte, err error) {
+	// Handle data URLs (base64 encoded)
+	if strings.HasPrefix(imageURL, "data:") {
+		parts := strings.SplitN(imageURL, ",", 2)
+		if len(parts) != 2 {
+			err = fmt.Errorf("invalid data URL format")
+			return
+		}
+		if ret, err = base64.StdEncoding.DecodeString(parts[1]); err != nil {
+			err = fmt.Errorf("failed to decode data URL: %w", err)
+		}
+		return
+	}
+
+	// Handle HTTP URLs with context
+	var req *http.Request
+	if req, err = http.NewRequestWithContext(ctx, http.MethodGet, imageURL, nil); err != nil {
+		return
+	}
+
+	var resp *http.Response
+	if resp, err = o.httpClient.Do(req); err != nil {
+		return
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode >= http.StatusBadRequest {
+		err = fmt.Errorf("failed to fetch image %s: %s", imageURL, resp.Status)
+		return
+	}
+
+	ret, err = io.ReadAll(resp.Body)
+	return
+}
+
 func (o *Client) NeedsRawMode(modelName string) bool {
-	ollamaPrefixes := []string{
+	ollamaSearchStrings := []string{
 		"llama3",
 		"llama2",
 		"mistral",
 	}
-	for _, prefix := range ollamaPrefixes {
-		if strings.HasPrefix(modelName, prefix) {
+	for _, searchString := range ollamaSearchStrings {
+		if strings.Contains(modelName, searchString) {
 			return true
 		}
 	}
--- a/internal/plugins/ai/openai/direct_models.go
+++ b/internal/plugins/ai/openai/direct_models.go
@@ -0,0 +1,120 @@
+package openai
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"time"
+
+	"github.com/danielmiessler/fabric/internal/i18n"
+	debuglog "github.com/danielmiessler/fabric/internal/log"
+)
+
+// modelResponse represents a minimal model returned by the API.
+// This mirrors the shape used by OpenAI-compatible providers that return
+// either an array of models or an object with a `data` field.
+type modelResponse struct {
+	ID string `json:"id"`
+}
+
+// errorResponseLimit defines the maximum length of error response bodies for truncation.
+const errorResponseLimit = 1024
+
+// maxResponseSize defines the maximum size of response bodies to prevent memory exhaustion.
+const maxResponseSize = 10 * 1024 * 1024 // 10MB
+
+// FetchModelsDirectly is used to fetch models directly from the API when the
+// standard OpenAI SDK method fails due to a nonstandard format. This is useful
+// for providers that return a direct array of models (e.g., GitHub Models) or
+// other OpenAI-compatible implementations.
+func FetchModelsDirectly(ctx context.Context, baseURL, apiKey, providerName string) ([]string, error) {
+	if ctx == nil {
+		ctx = context.Background()
+	}
+	if baseURL == "" {
+		return nil, fmt.Errorf(i18n.T("openai_api_base_url_not_configured"), providerName)
+	}
+
+	// Build the /models endpoint URL
+	fullURL, err := url.JoinPath(baseURL, "models")
+	if err != nil {
+		return nil, fmt.Errorf(i18n.T("openai_failed_to_create_models_url"), err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, fullURL, nil)
+	if err != nil {
+		return nil, err
+	}
+
+	req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", apiKey))
+	req.Header.Set("Accept", "application/json")
+
+	// TODO: Consider reusing a single http.Client instance (e.g., as a field on Client) instead of allocating a new one for
+	// each request.
+	client := &http.Client{
+		Timeout: 10 * time.Second,
+	}
+	resp, err := client.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+
+	if resp.StatusCode != http.StatusOK {
+		// Read the response body for debugging, but limit the number of bytes read
+		bodyBytes, readErr := io.ReadAll(io.LimitReader(resp.Body, errorResponseLimit))
+		if readErr != nil {
+			return nil, fmt.Errorf(i18n.T("openai_unexpected_status_code_read_error"),
+				resp.StatusCode, providerName, readErr)
+		}
+		bodyString := string(bodyBytes)
+		return nil, fmt.Errorf(i18n.T("openai_unexpected_status_code_with_body"),
+			resp.StatusCode, providerName, bodyString)
+	}
+
+	// Read the response body once, with a size limit to prevent memory exhaustion
+	// Read up to maxResponseSize + 1 bytes to detect truncation
+	bodyBytes, err := io.ReadAll(io.LimitReader(resp.Body, maxResponseSize+1))
+	if err != nil {
+		return nil, err
+	}
+	if len(bodyBytes) > maxResponseSize {
+		return nil, fmt.Errorf(i18n.T("openai_models_response_too_large"), providerName, maxResponseSize)
+	}
+
+	// Try to parse as an object with data field (OpenAI format)
+	var openAIFormat struct {
+		Data []modelResponse `json:"data"`
+	}
+	// Try to parse as a direct array
+	var directArray []modelResponse
+
+	if err := json.Unmarshal(bodyBytes, &openAIFormat); err == nil {
+		debuglog.Debug(debuglog.Detailed, "Successfully parsed models response from %s using OpenAI format (found %d models)\n", providerName, len(openAIFormat.Data))
+		return extractModelIDs(openAIFormat.Data), nil
+	}
+
+	if err := json.Unmarshal(bodyBytes, &directArray); err == nil {
+		debuglog.Debug(debuglog.Detailed, "Successfully parsed models response from %s using direct array format (found %d models)\n", providerName, len(directArray))
+		return extractModelIDs(directArray), nil
+	}
+
+	var truncatedBody string
+	if len(bodyBytes) > errorResponseLimit {
+		truncatedBody = string(bodyBytes[:errorResponseLimit]) + "..."
+	} else {
+		truncatedBody = string(bodyBytes)
+	}
+	return nil, fmt.Errorf(i18n.T("openai_unable_to_parse_models_response"), truncatedBody)
+}
+
+func extractModelIDs(models []modelResponse) []string {
+	modelIDs := make([]string, 0, len(models))
+	for _, model := range models {
+		modelIDs = append(modelIDs, model.ID)
+	}
+	return modelIDs
+}
--- a/internal/plugins/ai/openai/openai.go
+++ b/internal/plugins/ai/openai/openai.go
@@ -8,6 +8,7 @@ import (

 	"github.com/danielmiessler/fabric/internal/chat"
 	"github.com/danielmiessler/fabric/internal/domain"
+	debuglog "github.com/danielmiessler/fabric/internal/log"
 	"github.com/danielmiessler/fabric/internal/plugins"
 	openai "github.com/openai/openai-go"
 	"github.com/openai/openai-go/option"
@@ -83,13 +84,19 @@ func (o *Client) configure() (ret error) {

 func (o *Client) ListModels() (ret []string, err error) {
 	var page *pagination.Page[openai.Model]
-	if page, err = o.ApiClient.Models.List(context.Background()); err != nil {
-		return
+	if page, err = o.ApiClient.Models.List(context.Background()); err == nil {
+		for _, mod := range page.Data {
+			ret = append(ret, mod.ID)
+		}
+		// SDK succeeded - return the result even if empty
+		return ret, nil
 	}
-	for _, mod := range page.Data {
-		ret = append(ret, mod.ID)
-	}
-	return
+
+	// SDK returned an error - fall back to direct API fetch.
+	// Some providers (e.g., GitHub Models) return non-standard response formats
+	// that the SDK fails to parse.
+	debuglog.Debug(debuglog.Basic, "SDK Models.List failed for %s: %v, falling back to direct API fetch\n", o.GetName(), err)
+	return FetchModelsDirectly(context.Background(), o.ApiBaseURL.Value, o.ApiKey.Value, o.GetName())
 }

 func (o *Client) SendStream(
@@ -165,10 +172,11 @@ func (o *Client) supportsResponsesAPI() bool {

 func (o *Client) NeedsRawMode(modelName string) bool {
 	openaiModelsPrefixes := []string{
+		"glm",
+		"gpt-5",
 		"o1",
 		"o3",
 		"o4",
-		"gpt-5",
 	}
 	openAIModelsNeedingRaw := []string{
 		"gpt-4o-mini-search-preview",
--- a/internal/plugins/ai/openai/openai_image.go
+++ b/internal/plugins/ai/openai/openai_image.go
@@ -8,6 +8,7 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
+	"slices"
 	"strings"

 	"github.com/danielmiessler/fabric/internal/domain"
@@ -31,12 +32,7 @@ var ImageGenerationSupportedModels = []string{

 // supportsImageGeneration checks if the given model supports the image_generation tool
 func supportsImageGeneration(model string) bool {
-	for _, supportedModel := range ImageGenerationSupportedModels {
-		if model == supportedModel {
-			return true
-		}
-	}
-	return false
+	return slices.Contains(ImageGenerationSupportedModels, model)
 }

 // getOutputFormatFromExtension determines the API output format based on file extension
--- a/internal/plugins/ai/openai/openai_image_test.go
+++ b/internal/plugins/ai/openai/openai_image_test.go
@@ -345,7 +345,7 @@ func TestAddImageGenerationToolWithUserParameters(t *testing.T) {
 	tests := []struct {
 		name     string
 		opts     *domain.ChatOptions
-		expected map[string]interface{}
+		expected map[string]any
 	}{
 		{
 			name: "All parameters specified",
@@ -356,7 +356,7 @@ func TestAddImageGenerationToolWithUserParameters(t *testing.T) {
 				ImageBackground:  "transparent",
 				ImageCompression: 0, // Not applicable for PNG
 			},
-			expected: map[string]interface{}{
+			expected: map[string]any{
 				"size":          "1536x1024",
 				"quality":       "high",
 				"background":    "transparent",
@@ -372,7 +372,7 @@ func TestAddImageGenerationToolWithUserParameters(t *testing.T) {
 				ImageBackground:  "opaque",
 				ImageCompression: 75,
 			},
-			expected: map[string]interface{}{
+			expected: map[string]any{
 				"size":               "1024x1024",
 				"quality":            "medium",
 				"background":         "opaque",
@@ -386,7 +386,7 @@ func TestAddImageGenerationToolWithUserParameters(t *testing.T) {
 				ImageFile:    "/tmp/test.webp",
 				ImageQuality: "low",
 			},
-			expected: map[string]interface{}{
+			expected: map[string]any{
 				"quality":       "low",
 				"output_format": "webp",
 			},
@@ -396,7 +396,7 @@ func TestAddImageGenerationToolWithUserParameters(t *testing.T) {
 			opts: &domain.ChatOptions{
 				ImageFile: "/tmp/test.png",
 			},
-			expected: map[string]interface{}{
+			expected: map[string]any{
 				"output_format": "png",
 			},
 		},
--- a/internal/plugins/ai/openai/openai_models_test.go
+++ b/internal/plugins/ai/openai/openai_models_test.go
@@ -0,0 +1,58 @@
+package openai
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+// Ensures we can fetch models directly when a provider returns a direct array of models
+// instead of the standard OpenAI list response structure.
+func TestFetchModelsDirectly_DirectArray(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, "/models", r.URL.Path)
+		w.Header().Set("Content-Type", "application/json")
+		_, err := w.Write([]byte(`[{"id":"github-model"}]`))
+		assert.NoError(t, err)
+	}))
+	defer srv.Close()
+
+	models, err := FetchModelsDirectly(context.Background(), srv.URL, "test-key", "TestProvider")
+	assert.NoError(t, err)
+	assert.Equal(t, 1, len(models))
+	assert.Equal(t, "github-model", models[0])
+}
+
+// Ensures we can fetch models when a provider returns the standard OpenAI format
+func TestFetchModelsDirectly_OpenAIFormat(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, "/models", r.URL.Path)
+		w.Header().Set("Content-Type", "application/json")
+		_, err := w.Write([]byte(`{"data":[{"id":"openai-model"}]}`))
+		assert.NoError(t, err)
+	}))
+	defer srv.Close()
+
+	models, err := FetchModelsDirectly(context.Background(), srv.URL, "test-key", "TestProvider")
+	assert.NoError(t, err)
+	assert.Equal(t, 1, len(models))
+	assert.Equal(t, "openai-model", models[0])
+}
+
+// Ensures we handle empty model lists correctly
+func TestFetchModelsDirectly_EmptyArray(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, "/models", r.URL.Path)
+		w.Header().Set("Content-Type", "application/json")
+		_, err := w.Write([]byte(`[]`))
+		assert.NoError(t, err)
+	}))
+	defer srv.Close()
+
+	models, err := FetchModelsDirectly(context.Background(), srv.URL, "test-key", "TestProvider")
+	assert.NoError(t, err)
+	assert.Equal(t, 0, len(models))
+}
--- a/internal/plugins/ai/openai/openai_test.go
+++ b/internal/plugins/ai/openai/openai_test.go
@@ -16,7 +16,7 @@ func TestBuildResponseRequestWithMaxTokens(t *testing.T) {

 	var msgs []*chat.ChatCompletionMessage

-	for i := 0; i < 2; i++ {
+	for range 2 {
 		msgs = append(msgs, &chat.ChatCompletionMessage{
 			Role:    "User",
 			Content: "My msg",
@@ -42,7 +42,7 @@ func TestBuildResponseRequestNoMaxTokens(t *testing.T) {

 	var msgs []*chat.ChatCompletionMessage

-	for i := 0; i < 2; i++ {
+	for range 2 {
 		msgs = append(msgs, &chat.ChatCompletionMessage{
 			Role:    "User",
 			Content: "My msg",
--- a/internal/plugins/ai/openai_compatible/direct_models_call.go
+++ b/internal/plugins/ai/openai_compatible/direct_models_call.go
@@ -2,104 +2,12 @@ package openai_compatible

 import (
 	"context"
-	"encoding/json"
-	"fmt"
-	"io"
-	"net/http"
-	"net/url"
-	"time"
+
+	"github.com/danielmiessler/fabric/internal/plugins/ai/openai"
 )

-// Model represents a model returned by the API
-type Model struct {
-	ID string `json:"id"`
-}
-
-// ErrorResponseLimit defines the maximum length of error response bodies for truncation.
-const errorResponseLimit = 1024 // Limit for error response body size
-
-// DirectlyGetModels is used to fetch models directly from the API
-// when the standard OpenAI SDK method fails due to a nonstandard format.
-// This is useful for providers like Together that return a direct array of models.
+// DirectlyGetModels is used to fetch models directly from the API when the
+// standard OpenAI SDK method fails due to a nonstandard format.
 func (c *Client) DirectlyGetModels(ctx context.Context) ([]string, error) {
-	if ctx == nil {
-		ctx = context.Background()
-	}
-	baseURL := c.ApiBaseURL.Value
-	if baseURL == "" {
-		return nil, fmt.Errorf("API base URL not configured for provider %s", c.GetName())
-	}
-
-	// Build the /models endpoint URL
-	fullURL, err := url.JoinPath(baseURL, "models")
-	if err != nil {
-		return nil, fmt.Errorf("failed to create models URL: %w", err)
-	}
-
-	req, err := http.NewRequestWithContext(ctx, "GET", fullURL, nil)
-	if err != nil {
-		return nil, err
-	}
-
-	req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", c.ApiKey.Value))
-	req.Header.Set("Accept", "application/json")
-
-	// TODO: Consider reusing a single http.Client instance (e.g., as a field on Client) instead of allocating a new one for each request.
-
-	client := &http.Client{
-		Timeout: 10 * time.Second,
-	}
-	resp, err := client.Do(req)
-	if err != nil {
-		return nil, err
-	}
-	defer resp.Body.Close()
-
-	if resp.StatusCode != http.StatusOK {
-		// Read the response body for debugging
-		bodyBytes, _ := io.ReadAll(resp.Body)
-		bodyString := string(bodyBytes)
-		if len(bodyString) > errorResponseLimit { // Truncate if too large
-			bodyString = bodyString[:errorResponseLimit] + "..."
-		}
-		return nil, fmt.Errorf("unexpected status code: %d from provider %s, response body: %s",
-			resp.StatusCode, c.GetName(), bodyString)
-	}
-
-	// Read the response body once
-	bodyBytes, err := io.ReadAll(resp.Body)
-	if err != nil {
-		return nil, err
-	}
-
-	// Try to parse as an object with data field (OpenAI format)
-	var openAIFormat struct {
-		Data []Model `json:"data"`
-	}
-	// Try to parse as a direct array (Together format)
-	var directArray []Model
-
-	if err := json.Unmarshal(bodyBytes, &openAIFormat); err == nil && len(openAIFormat.Data) > 0 {
-		return extractModelIDs(openAIFormat.Data), nil
-	}
-
-	if err := json.Unmarshal(bodyBytes, &directArray); err == nil && len(directArray) > 0 {
-		return extractModelIDs(directArray), nil
-	}
-
-	var truncatedBody string
-	if len(bodyBytes) > errorResponseLimit {
-		truncatedBody = string(bodyBytes[:errorResponseLimit]) + "..."
-	} else {
-		truncatedBody = string(bodyBytes)
-	}
-	return nil, fmt.Errorf("unable to parse models response; raw response: %s", truncatedBody)
-}
-
-func extractModelIDs(models []Model) []string {
-	modelIDs := make([]string, 0, len(models))
-	for _, model := range models {
-		modelIDs = append(modelIDs, model.ID)
-	}
-	return modelIDs
+	return openai.FetchModelsDirectly(ctx, c.ApiBaseURL.Value, c.ApiKey.Value, c.GetName())
 }
--- a/internal/plugins/ai/openai_compatible/providers_config.go
+++ b/internal/plugins/ai/openai_compatible/providers_config.go
@@ -2,6 +2,7 @@ package openai_compatible

 import (
 	"context"
+	"fmt"
 	"os"
 	"strings"

@@ -12,17 +13,21 @@ import (
 type ProviderConfig struct {
 	Name                string
 	BaseURL             string
-	ImplementsResponses bool // Whether the provider supports OpenAI's new Responses API
+	ModelsURL           string // Optional: Custom endpoint for listing models (if different from BaseURL/models)
+	ImplementsResponses bool   // Whether the provider supports OpenAI's new Responses API
 }

 // Client is the common structure for all OpenAI-compatible providers
 type Client struct {
 	*openai.Client
+	modelsURL string // Custom URL for listing models (if different from BaseURL/models)
 }

 // NewClient creates a new OpenAI-compatible client for the specified provider
 func NewClient(providerConfig ProviderConfig) *Client {
-	client := &Client{}
+	client := &Client{
+		modelsURL: providerConfig.ModelsURL,
+	}
 	client.Client = openai.NewClientCompatibleWithResponses(
 		providerConfig.Name,
 		providerConfig.BaseURL,
@@ -34,17 +39,89 @@ func NewClient(providerConfig ProviderConfig) *Client {

 // ListModels overrides the default ListModels to handle different response formats
 func (c *Client) ListModels() ([]string, error) {
+	// If a custom models URL is provided, handle it
+	if c.modelsURL != "" {
+		// Check for static model list
+		if strings.HasPrefix(c.modelsURL, "static:") {
+			return c.getStaticModels(c.modelsURL)
+		}
+		// TODO: Handle context properly in Fabric by accepting and propagating a context.Context
+		// instead of creating a new one here.
+		return openai.FetchModelsDirectly(context.Background(), c.modelsURL, c.Client.ApiKey.Value, c.GetName())
+	}
+
 	// First try the standard OpenAI SDK approach
 	models, err := c.Client.ListModels()
 	if err == nil && len(models) > 0 { // only return if OpenAI SDK returns models
 		return models, nil
 	}

-	// TODO: Handle context properly in Fabric by accepting and propagating a context.Context
-	// instead of creating a new one here.
+	// Fall back to direct API fetch
 	return c.DirectlyGetModels(context.Background())
 }

+// getStaticModels returns a predefined list of models for providers that don't support model discovery
+func (c *Client) getStaticModels(modelsKey string) ([]string, error) {
+	switch modelsKey {
+	case "static:abacus":
+		return []string{
+			"route-llm",
+			"gpt-4o-2024-11-20",
+			"gpt-4o-mini",
+			"o4-mini",
+			"o3-pro",
+			"o3",
+			"o3-mini",
+			"gpt-4.1",
+			"gpt-4.1-mini",
+			"gpt-4.1-nano",
+			"gpt-5",
+			"gpt-5-mini",
+			"gpt-5-nano",
+			"gpt-5.1",
+			"gpt-5.1-chat-latest",
+			"openai/gpt-oss-120b",
+			"claude-3-7-sonnet-20250219",
+			"claude-sonnet-4-20250514",
+			"claude-opus-4-20250514",
+			"claude-opus-4-1-20250805",
+			"claude-sonnet-4-5-20250929",
+			"claude-haiku-4-5-20251001",
+			"claude-opus-4-5-20251101",
+			"meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
+			"meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo",
+			"meta-llama/Meta-Llama-3.1-70B-Instruct",
+			"meta-llama/Meta-Llama-3.1-8B-Instruct",
+			"llama-3.3-70b-versatile",
+			"gemini-2.0-flash-001",
+			"gemini-2.0-pro-exp-02-05",
+			"gemini-2.5-pro",
+			"gemini-2.5-flash",
+			"gemini-3-pro-preview",
+			"qwen-2.5-coder-32b",
+			"Qwen/Qwen2.5-72B-Instruct",
+			"Qwen/QwQ-32B",
+			"Qwen/Qwen3-235B-A22B-Instruct-2507",
+			"Qwen/Qwen3-32B",
+			"qwen/qwen3-coder-480b-a35b-instruct",
+			"qwen/qwen3-Max",
+			"grok-4-0709",
+			"grok-4-fast-non-reasoning",
+			"grok-4-1-fast-non-reasoning",
+			"grok-code-fast-1",
+			"kimi-k2-turbo-preview",
+			"deepseek/deepseek-v3.1",
+			"deepseek-ai/DeepSeek-V3.1-Terminus",
+			"deepseek-ai/DeepSeek-R1",
+			"deepseek-ai/DeepSeek-V3.2",
+			"zai-org/glm-4.5",
+			"zai-org/glm-4.6",
+		}, nil
+	default:
+		return nil, fmt.Errorf("unknown static model list: %s", modelsKey)
+	}
+}
+
 // ProviderMap is a map of provider name to ProviderConfig for O(1) lookup
 var ProviderMap = map[string]ProviderConfig{
 	"AIML": {
@@ -62,6 +139,12 @@ var ProviderMap = map[string]ProviderConfig{
 		BaseURL:             "https://api.deepseek.com",
 		ImplementsResponses: false,
 	},
+	"GitHub": {
+		Name:                "GitHub",
+		BaseURL:             "https://models.github.ai/inference",
+		ModelsURL:           "https://models.github.ai/catalog", // FetchModelsDirectly will append /models
+		ImplementsResponses: false,
+	},
 	"GrokAI": {
 		Name:                "GrokAI",
 		BaseURL:             "https://api.x.ai/v1",
@@ -107,6 +190,17 @@ var ProviderMap = map[string]ProviderConfig{
 		BaseURL:             "https://api.venice.ai/api/v1",
 		ImplementsResponses: false,
 	},
+	"Z AI": {
+		Name:                "Z AI",
+		BaseURL:             "https://api.z.ai/api/paas/v4",
+		ImplementsResponses: false,
+	},
+	"Abacus": {
+		Name:                "Abacus",
+		BaseURL:             "https://routellm.abacus.ai/v1/",
+		ModelsURL:           "static:abacus", // Special marker for static model list
+		ImplementsResponses: false,
+	},
 }

 // GetProviderByName returns the provider configuration for a given name with O(1) lookup
--- a/internal/plugins/ai/openai_compatible/providers_config_test.go
+++ b/internal/plugins/ai/openai_compatible/providers_config_test.go
@@ -20,6 +20,16 @@ func TestCreateClient(t *testing.T) {
 			provider: "Groq",
 			exists:   true,
 		},
+		{
+			name:     "Existing provider - Z AI",
+			provider: "Z AI",
+			exists:   true,
+		},
+		{
+			name:     "Existing provider - Abacus",
+			provider: "Abacus",
+			exists:   true,
+		},
 		{
 			name:     "Non-existent provider",
 			provider: "NonExistent",
--- a/internal/plugins/ai/perplexity/perplexity.go
+++ b/internal/plugins/ai/perplexity/perplexity.go
@@ -4,6 +4,7 @@ import (
 	"context"
 	"fmt"
 	"os"
+	"strings"
 	"sync"

 	"github.com/danielmiessler/fabric/internal/domain"
@@ -107,18 +108,19 @@ func (c *Client) Send(ctx context.Context, msgs []*chat.ChatCompletionMessage, o
 		return "", fmt.Errorf("perplexity API request failed: %w", err) // Corrected capitalization
 	}

-	content := resp.GetLastContent()
+	var content strings.Builder
+	content.WriteString(resp.GetLastContent())

 	// Append citations if available
 	citations := resp.GetCitations()
 	if len(citations) > 0 {
-		content += "\n\n# CITATIONS\n\n"
+		content.WriteString("\n\n# CITATIONS\n\n")
 		for i, citation := range citations {
-			content += fmt.Sprintf("- [%d] %s\n", i+1, citation)
+			content.WriteString(fmt.Sprintf("- [%d] %s\n", i+1, citation))
 		}
 	}

-	return content, nil
+	return content.String(), nil
 }

 func (c *Client) SendStream(msgs []*chat.ChatCompletionMessage, opts *domain.ChatOptions, channel chan string) error {
--- a/internal/plugins/ai/vendors.go
+++ b/internal/plugins/ai/vendors.go
@@ -25,9 +25,12 @@ type VendorsManager struct {
 	Models        *VendorsModels
 }

+// AddVendors registers one or more vendors with the manager.
+// Vendors are stored with lowercase keys to enable case-insensitive lookup.
 func (o *VendorsManager) AddVendors(vendors ...Vendor) {
 	for _, vendor := range vendors {
-		o.VendorsByName[vendor.GetName()] = vendor
+		name := strings.ToLower(vendor.GetName())
+		o.VendorsByName[name] = vendor
 		o.Vendors = append(o.Vendors, vendor)
 	}
 }
@@ -63,8 +66,10 @@ func (o *VendorsManager) HasVendors() bool {
 	return len(o.Vendors) > 0
 }

+// FindByName returns a vendor by name. Lookup is case-insensitive.
+// For example, "OpenAI", "openai", and "OPENAI" all match the same vendor.
 func (o *VendorsManager) FindByName(name string) Vendor {
-	return o.VendorsByName[name]
+	return o.VendorsByName[strings.ToLower(name)]
 }

 func (o *VendorsManager) readModels() (err error) {
@@ -143,9 +148,9 @@ func (o *VendorsManager) SetupVendor(vendorName string, configuredVendors map[st
 func (o *VendorsManager) setupVendorTo(vendor Vendor, configuredVendors map[string]Vendor) {
 	if vendorErr := vendor.Setup(); vendorErr == nil {
 		fmt.Printf("[%v] configured\n", vendor.GetName())
-		configuredVendors[vendor.GetName()] = vendor
+		configuredVendors[strings.ToLower(vendor.GetName())] = vendor
 	} else {
-		delete(configuredVendors, vendor.GetName())
+		delete(configuredVendors, strings.ToLower(vendor.GetName()))
 		fmt.Printf("[%v] skipped\n", vendor.GetName())
 	}
 }
--- a/internal/plugins/ai/vendors_test.go
+++ b/internal/plugins/ai/vendors_test.go
@@ -0,0 +1,66 @@
+package ai
+
+import (
+	"bytes"
+	"context"
+	"testing"
+
+	"github.com/danielmiessler/fabric/internal/chat"
+	"github.com/danielmiessler/fabric/internal/domain"
+)
+
+type stubVendor struct {
+	name string
+}
+
+func (v *stubVendor) GetName() string                       { return v.name }
+func (v *stubVendor) GetSetupDescription() string           { return "" }
+func (v *stubVendor) IsConfigured() bool                    { return true }
+func (v *stubVendor) Configure() error                      { return nil }
+func (v *stubVendor) Setup() error                          { return nil }
+func (v *stubVendor) SetupFillEnvFileContent(*bytes.Buffer) {}
+func (v *stubVendor) ListModels() ([]string, error)         { return nil, nil }
+func (v *stubVendor) SendStream([]*chat.ChatCompletionMessage, *domain.ChatOptions, chan string) error {
+	return nil
+}
+func (v *stubVendor) Send(context.Context, []*chat.ChatCompletionMessage, *domain.ChatOptions) (string, error) {
+	return "", nil
+}
+func (v *stubVendor) NeedsRawMode(string) bool { return false }
+
+func TestVendorsManagerFindByNameCaseInsensitive(t *testing.T) {
+	manager := NewVendorsManager()
+	vendor := &stubVendor{name: "OpenAI"}
+
+	manager.AddVendors(vendor)
+
+	if got := manager.FindByName("openai"); got != vendor {
+		t.Fatalf("FindByName lowercase = %v, want %v", got, vendor)
+	}
+
+	if got := manager.FindByName("OPENAI"); got != vendor {
+		t.Fatalf("FindByName uppercase = %v, want %v", got, vendor)
+	}
+
+	if got := manager.FindByName("OpenAI"); got != vendor {
+		t.Fatalf("FindByName mixed case = %v, want %v", got, vendor)
+	}
+}
+
+func TestVendorsManagerSetupVendorToCaseInsensitive(t *testing.T) {
+	manager := NewVendorsManager()
+	vendor := &stubVendor{name: "OpenAI"}
+
+	configured := map[string]Vendor{}
+	manager.setupVendorTo(vendor, configured)
+
+	// Verify vendor is stored with lowercase key
+	if _, ok := configured["openai"]; !ok {
+		t.Fatalf("setupVendorTo should store vendor using lowercase key")
+	}
+
+	// Verify original case key is not used
+	if _, ok := configured["OpenAI"]; ok {
+		t.Fatalf("setupVendorTo should not store vendor using original case key")
+	}
+}
--- a/internal/plugins/db/fsdb/patterns.go
+++ b/internal/plugins/db/fsdb/patterns.go
@@ -11,8 +11,6 @@ import (
 	"github.com/danielmiessler/fabric/internal/util"
 )

-const inputSentinel = "__FABRIC_INPUT_SENTINEL_TOKEN__"
-
 type PatternsEntity struct {
 	*StorageEntity
 	SystemPatternFile      string
@@ -96,18 +94,18 @@ func (o *PatternsEntity) applyVariables(

 	// Temporarily replace {{input}} with a sentinel token to protect it
 	// from recursive variable resolution
-	withSentinel := strings.ReplaceAll(pattern.Pattern, "{{input}}", inputSentinel)
+	withSentinel := strings.ReplaceAll(pattern.Pattern, "{{input}}", template.InputSentinel)

 	// Process all other template variables in the pattern
-	// At this point, our sentinel ensures {{input}} won't be affected
+	// Pass the actual input so extension calls can use {{input}} within their value parameter
 	var processed string
-	if processed, err = template.ApplyTemplate(withSentinel, variables, ""); err != nil {
+	if processed, err = template.ApplyTemplate(withSentinel, variables, input); err != nil {
 		return
 	}

 	// Finally, replace our sentinel with the actual user input
 	// The input has already been processed for variables if InputHasVars was true
-	pattern.Pattern = strings.ReplaceAll(processed, inputSentinel, input)
+	pattern.Pattern = strings.ReplaceAll(processed, template.InputSentinel, input)
 	return
 }

--- a/internal/plugins/db/fsdb/storage.go
+++ b/internal/plugins/db/fsdb/storage.go
@@ -134,7 +134,7 @@ func (o *StorageEntity) buildFileName(name string) string {
 	return fmt.Sprintf("%s%v", name, o.FileExtension)
 }

-func (o *StorageEntity) SaveAsJson(name string, item interface{}) (err error) {
+func (o *StorageEntity) SaveAsJson(name string, item any) (err error) {
 	var jsonString []byte
 	if jsonString, err = json.Marshal(item); err == nil {
 		err = o.Save(name, jsonString)
@@ -145,7 +145,7 @@ func (o *StorageEntity) SaveAsJson(name string, item interface{}) (err error) {
 	return err
 }

-func (o *StorageEntity) LoadAsJson(name string, item interface{}) (err error) {
+func (o *StorageEntity) LoadAsJson(name string, item any) (err error) {
 	var content []byte
 	if content, err = o.Load(name); err != nil {
 		return
--- a/internal/plugins/plugin.go
+++ b/internal/plugins/plugin.go
@@ -92,7 +92,11 @@ func (o *PluginBase) Setup() (err error) {
 		return
 	}

-	err = o.Configure()
+	// After Setup, run ConfigureCustom if present, but skip re-validation
+	// since Ask() already validated user input (or allowed explicit reset)
+	if o.ConfigureCustom != nil {
+		err = o.ConfigureCustom()
+	}
 	return
 }

@@ -198,16 +202,21 @@ func (o *SetupQuestion) Ask(label string) (err error) {
 	var answer string
 	fmt.Scanln(&answer)
 	answer = strings.TrimRight(answer, "\n")
+	isReset := strings.ToLower(answer) == AnswerReset
 	if answer == "" {
 		answer = o.Value
-	} else if strings.ToLower(answer) == AnswerReset {
+	} else if isReset {
 		answer = ""
 	}
-	err = o.OnAnswer(answer)
+	err = o.OnAnswerWithReset(answer, isReset)
 	return
 }

 func (o *SetupQuestion) OnAnswer(answer string) (err error) {
+	return o.OnAnswerWithReset(answer, false)
+}
+
+func (o *SetupQuestion) OnAnswerWithReset(answer string, isReset bool) (err error) {
 	if o.Type == SettingTypeBool {
 		if answer == "" {
 			o.Value = ""
@@ -226,6 +235,11 @@ func (o *SetupQuestion) OnAnswer(answer string) (err error) {
 			return
 		}
 	}
+	// Skip validation when explicitly resetting a value - the user intentionally
+	// wants to clear the value even if it's required
+	if isReset {
+		return nil
+	}
 	err = o.IsValidErr()
 	return
 }
--- a/internal/plugins/plugin_test.go
+++ b/internal/plugins/plugin_test.go
@@ -116,6 +116,91 @@ func TestSetupQuestion_Ask(t *testing.T) {
 	assert.Equal(t, "user_value", setting.Value)
 }

+func TestSetupQuestion_Ask_Reset(t *testing.T) {
+	// Test that resetting a required field doesn't produce an error
+	setting := &Setting{
+		EnvVariable: "TEST_RESET_SETTING",
+		Value:       "existing_value",
+		Required:    true,
+	}
+	question := &SetupQuestion{
+		Setting:  setting,
+		Question: "Enter test setting:",
+	}
+	input := "reset\n"
+	fmtInput := captureInput(input)
+	defer fmtInput()
+	err := question.Ask("TestConfigurable")
+	// Should NOT return an error even though the field is required
+	assert.NoError(t, err)
+	// Value should be cleared
+	assert.Equal(t, "", setting.Value)
+}
+
+func TestSetupQuestion_OnAnswerWithReset(t *testing.T) {
+	tests := []struct {
+		name        string
+		setting     *Setting
+		answer      string
+		isReset     bool
+		expectError bool
+		expectValue string
+	}{
+		{
+			name: "reset required field should not error",
+			setting: &Setting{
+				EnvVariable: "TEST_SETTING",
+				Value:       "old_value",
+				Required:    true,
+			},
+			answer:      "",
+			isReset:     true,
+			expectError: false,
+			expectValue: "",
+		},
+		{
+			name: "empty answer on required field should error",
+			setting: &Setting{
+				EnvVariable: "TEST_SETTING",
+				Value:       "",
+				Required:    true,
+			},
+			answer:      "",
+			isReset:     false,
+			expectError: true,
+			expectValue: "",
+		},
+		{
+			name: "valid answer on required field should not error",
+			setting: &Setting{
+				EnvVariable: "TEST_SETTING",
+				Value:       "",
+				Required:    true,
+			},
+			answer:      "new_value",
+			isReset:     false,
+			expectError: false,
+			expectValue: "new_value",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			question := &SetupQuestion{
+				Setting:  tt.setting,
+				Question: "Test question",
+			}
+			err := question.OnAnswerWithReset(tt.answer, tt.isReset)
+			if tt.expectError {
+				assert.Error(t, err)
+			} else {
+				assert.NoError(t, err)
+			}
+			assert.Equal(t, tt.expectValue, tt.setting.Value)
+		})
+	}
+}
+
 func TestSettings_IsConfigured(t *testing.T) {
 	settings := Settings{
 		{EnvVariable: "TEST_SETTING1", Value: "value1", Required: true},
--- a/internal/plugins/template/Examples/README.md
+++ b/internal/plugins/template/Examples/README.md
@@ -1,9 +1,24 @@

 # Fabric Extensions: Complete Guide

+## Important: Extensions Only Work in Patterns
+
+**Extensions are ONLY processed when used within pattern files, not via direct piping to fabric.**
+
+```bash
+# ❌ This DOES NOT WORK - extensions are not processed in stdin
+echo "{{ext:word-generator:generate:3}}" | fabric
+
+# ✅ This WORKS - extensions are processed within patterns
+fabric -p my-pattern-with-extensions.md
+```
+
+When you pipe directly to fabric without a pattern, the input goes straight to the LLM without template processing. Extensions are only evaluated during pattern template processing via `ApplyTemplate()`.
+
 ## Understanding Extension Architecture

 ### Registry Structure
+
 The extension registry is stored at `~/.config/fabric/extensions/extensions.yaml` and tracks registered extensions:

 ```yaml
@@ -17,6 +32,7 @@ extensions:
 The registry maintains security through hash verification of both configs and executables.

 ### Extension Configuration
+
 Each extension requires a YAML configuration file with the following structure:

 ```yaml
@@ -42,8 +58,10 @@ config:                      # Output configuration
 ```

 ### Directory Structure
+
 Recommended organization:
-```
+
+```text
 ~/.config/fabric/extensions/
 ├── bin/           # Extension executables
 ├── configs/       # Extension YAML configs
@@ -51,9 +69,11 @@ Recommended organization:
 ```

 ## Example 1: Python Wrapper (Word Generator)
+
 A simple example wrapping a Python script.

 ### 1. Position Files
+
 ```bash
 # Create directories
 mkdir -p ~/.config/fabric/extensions/{bin,configs}
@@ -64,7 +84,9 @@ chmod +x ~/.config/fabric/extensions/bin/word-generator.py
 ```

 ### 2. Configure
+
 Create `~/.config/fabric/extensions/configs/word-generator.yaml`:
+
 ```yaml
 name: word-generator
 executable: "~/.config/fabric/extensions/bin/word-generator.py"
@@ -83,22 +105,26 @@ config:
 ```

 ### 3. Register & Run
+
 ```bash
 # Register
 fabric --addextension ~/.config/fabric/extensions/configs/word-generator.yaml

-# Run (generate 3 random words)
-echo "{{ext:word-generator:generate:3}}" | fabric
+# Extensions must be used within patterns (see "Extensions in patterns" section below)
+# Direct piping to fabric will NOT process extension syntax
 ```

 ## Example 2: Direct Executable (SQLite3)
+
 Using a system executable directly.

 copy the memories to your home directory
 ~/memories.db

 ### 1. Configure
+
 Create `~/.config/fabric/extensions/configs/memory-query.yaml`:
+
 ```yaml
 name: memory-query
 executable: "/usr/bin/sqlite3"
@@ -123,19 +149,19 @@ config:
 ```

 ### 2. Register & Run
+
 ```bash
 # Register
 fabric --addextension ~/.config/fabric/extensions/configs/memory-query.yaml

-# Run queries
-echo  "{{ext:memory-query:all}}" | fabric
-echo  "{{ext:memory-query:byid:3}}" | fabric
+# Extensions must be used within patterns (see "Extensions in patterns" section below)
+# Direct piping to fabric will NOT process extension syntax
 ```

-
 ## Extension Management Commands

 ### Add Extension
+
 ```bash
 fabric --addextension ~/.config/fabric/extensions/configs/memory-query.yaml
 ```
@@ -143,25 +169,29 @@ fabric --addextension ~/.config/fabric/extensions/configs/memory-query.yaml
 Note : if the executable or config file changes, you must re-add the extension.
 This will recompute the hash for the extension.

-
 ### List Extensions
+
 ```bash
 fabric --listextensions
 ```
+
 Shows all registered extensions with their status and configuration details.

 ### Remove Extension
+
 ```bash
 fabric --rmextension <extension-name>
 ```
-Removes an extension from the registry.

+Removes an extension from the registry.

 ## Extensions in patterns

-```
-Create a pattern that use multiple extensions.
+**IMPORTANT**: Extensions are ONLY processed when used within pattern files, not via direct piping to fabric.

+Create a pattern file (e.g., `test_pattern.md`):
+
+```markdown
 These are my favorite
 {{ext:word-generator:generate:3}}

@@ -171,8 +201,30 @@ These are my least favorite
 what does this say about me?
 ```

+Run the pattern:
+
 ```bash
-./fabric -p ./plugins/template/Examples/test_pattern.md
+fabric -p ./internal/plugins/template/Examples/test_pattern.md
+```
+
+## Passing {{input}} to extensions inside patterns
+
+```text
+Create a pattern called ai_summarize that uses extensions (see openai.yaml and copy for claude)
+
+Summarize the responses from both AI models:
+
+OpenAI Response:
+{{ext:openai:chat:{{input}}}}
+
+Claude Response:
+{{ext:claude:chat:{{input}}}}
+
+```
+
+```bash
+echo "What is Artificial Intelligence" | ../fabric-fix -p ai_summarize
+
 ```

 ## Security Considerations
@@ -197,6 +249,7 @@ what does this say about me?
 ## Troubleshooting

 ### Common Issues
+
 1. **Registration Failures**
   - Verify file permissions
   - Check executable paths
@@ -214,10 +267,10 @@ what does this say about me?
   - Monitor disk space for file operations

 ### Debug Tips
+
 1. Enable verbose logging when available
 2. Check system logs for execution errors
 3. Verify extension dependencies
 4. Test extensions with minimal configurations first

-
-Would you like me to expand on any particular section or add more examples?
+Would you like me to expand on any particular section or add more examples?
--- a/internal/plugins/template/Examples/openai-chat.sh
+++ b/internal/plugins/template/Examples/openai-chat.sh
@@ -0,0 +1,20 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+INPUT=$(jq -R -s '.' <<< "$*")
+RESPONSE=$(curl "$OPENAI_API_BASE_URL/chat/completions" \
+  -s -w "\n%{http_code}" \
+  -H "Content-Type: application/json" \
+  -H "Authorization: Bearer $OPENAI_API_KEY" \
+  -d "{\"model\":\"gpt-4o-mini\",\"messages\":[{\"role\":\"user\",\"content\":$INPUT}]}")
+
+HTTP_CODE=$(echo "$RESPONSE" | tail -n1)
+BODY=$(echo "$RESPONSE" | sed '$d')
+
+if [[ "$HTTP_CODE" -ne 200 ]]; then
+    echo "Error: HTTP $HTTP_CODE" >&2
+    echo "$BODY" | jq -r '.error.message // "Unknown error"' >&2
+    exit 1
+fi
+
+echo "$BODY" | jq -r '.choices[0].message.content'
--- a/internal/plugins/template/Examples/openai.yaml
+++ b/internal/plugins/template/Examples/openai.yaml
@@ -0,0 +1,14 @@
+name: openai
+executable: "/path/to/your/openai-chat.sh"
+type: executable
+timeout: "30s"
+description: "Call OpenAI Chat Completions API"
+version: "1.0.0"
+
+operations:
+  chat:
+    cmd_template: "{{executable}} {{value}}"
+
+config:
+  output:
+    method: stdout
--- a/internal/plugins/template/constants.go
+++ b/internal/plugins/template/constants.go
@@ -0,0 +1,5 @@
+package template
+
+// InputSentinel is used to temporarily replace {{input}} during template processing
+// to prevent recursive variable resolution
+const InputSentinel = "__FABRIC_INPUT_SENTINEL_TOKEN__"
--- a/internal/plugins/template/extension_executor_test.go
+++ b/internal/plugins/template/extension_executor_test.go
@@ -187,9 +187,10 @@ esac`
 	executor := NewExtensionExecutor(registry)

 	// Helper function to create and register extension
-	createExtension := func(name, opName, cmdTemplate string, config map[string]interface{}) error {
+	createExtension := func(name, opName, cmdTemplate string, config map[string]any) error {
 		configPath := filepath.Join(tmpDir, name+".yaml")
-		configContent := `name: ` + name + `
+		var configContent strings.Builder
+		configContent.WriteString(`name: ` + name + `
 executable: ` + testScript + `
 type: executable
 timeout: 30s
@@ -199,14 +200,14 @@ operations:
 config:
  output:
    method: file
-    file_config:`
+    file_config:`)

 		// Add config options
 		for k, v := range config {
-			configContent += "\n      " + k + ": " + strings.TrimSpace(v.(string))
+			configContent.WriteString("\n      " + k + ": " + strings.TrimSpace(v.(string)))
 		}

-		if err := os.WriteFile(configPath, []byte(configContent), 0644); err != nil {
+		if err := os.WriteFile(configPath, []byte(configContent.String()), 0644); err != nil {
 			return err
 		}

@@ -216,7 +217,7 @@ config:
 	// Test basic fixed file output
 	t.Run("BasicFixedFile", func(t *testing.T) {
 		outputFile := filepath.Join(tmpDir, "output.txt")
-		config := map[string]interface{}{
+		config := map[string]any{
 			"output_file": `"output.txt"`,
 			"work_dir":    `"` + tmpDir + `"`,
 			"cleanup":     "true",
@@ -241,7 +242,7 @@ config:

 	// Test no work_dir specified
 	t.Run("NoWorkDir", func(t *testing.T) {
-		config := map[string]interface{}{
+		config := map[string]any{
 			"output_file": `"direct-output.txt"`,
 			"cleanup":     "true",
 		}
@@ -263,7 +264,7 @@ config:
 		outputFile := filepath.Join(tmpDir, "cleanup-test.txt")

 		// Test with cleanup enabled
-		config := map[string]interface{}{
+		config := map[string]any{
 			"output_file": `"cleanup-test.txt"`,
 			"work_dir":    `"` + tmpDir + `"`,
 			"cleanup":     "true",
@@ -307,7 +308,7 @@ config:
 	// Test error cases
 	t.Run("ErrorCases", func(t *testing.T) {
 		outputFile := filepath.Join(tmpDir, "error-test.txt")
-		config := map[string]interface{}{
+		config := map[string]any{
 			"output_file": `"error-test.txt"`,
 			"work_dir":    `"` + tmpDir + `"`,
 			"cleanup":     "true",
@@ -341,7 +342,7 @@ config:

 	// Test with missing output_file
 	t.Run("MissingOutputFile", func(t *testing.T) {
-		config := map[string]interface{}{
+		config := map[string]any{
 			"work_dir": `"` + tmpDir + `"`,
 			"cleanup":  "true",
 		}
--- a/internal/plugins/template/extension_registry.go
+++ b/internal/plugins/template/extension_registry.go
@@ -30,7 +30,7 @@ type ExtensionDefinition struct {
 	Operations map[string]OperationConfig `yaml:"operations"`

 	// Additional config
-	Config map[string]interface{} `yaml:"config"`
+	Config map[string]any `yaml:"config"`
 }

 type OperationConfig struct {
@@ -53,7 +53,7 @@ type ExtensionRegistry struct {

 // Helper methods for Config access
 func (e *ExtensionDefinition) GetOutputMethod() string {
-	if output, ok := e.Config["output"].(map[string]interface{}); ok {
+	if output, ok := e.Config["output"].(map[string]any); ok {
 		if method, ok := output["method"].(string); ok {
 			return method
 		}
@@ -61,9 +61,9 @@ func (e *ExtensionDefinition) GetOutputMethod() string {
 	return "stdout" // default to stdout if not specified
 }

-func (e *ExtensionDefinition) GetFileConfig() map[string]interface{} {
-	if output, ok := e.Config["output"].(map[string]interface{}); ok {
-		if fileConfig, ok := output["file_config"].(map[string]interface{}); ok {
+func (e *ExtensionDefinition) GetFileConfig() map[string]any {
+	if output, ok := e.Config["output"].(map[string]any); ok {
+		if fileConfig, ok := output["file_config"].(map[string]any); ok {
 			return fileConfig
 		}
 	}
@@ -140,6 +140,11 @@ func (r *ExtensionRegistry) Register(configPath string) error {
 		return fmt.Errorf("failed to hash executable: %w", err)
 	}

+	// Validate full extension definition (ensures operations and cmd_template present)
+	if err := r.validateExtensionDefinition(&ext); err != nil {
+		return fmt.Errorf("invalid extension definition: %w", err)
+	}
+
 	// Store entry
 	r.registry.Extensions[ext.Name] = &RegistryEntry{
 		ConfigPath:     absPath,
--- a/internal/plugins/template/template.go
+++ b/internal/plugins/template/template.go
@@ -33,156 +33,69 @@ func init() {
 var pluginPattern = regexp.MustCompile(`\{\{plugin:([^:]+):([^:]+)(?::([^}]+))?\}\}`)
 var extensionPattern = regexp.MustCompile(`\{\{ext:([^:]+):([^:]+)(?::([^}]+))?\}\}`)

-func debugf(format string, a ...interface{}) {
+func debugf(format string, a ...any) {
 	debuglog.Debug(debuglog.Trace, format, a...)
 }

-func ApplyTemplate(content string, variables map[string]string, input string) (string, error) {
-
-	var missingVars []string
-	r := regexp.MustCompile(`\{\{([^{}]+)\}\}`)
-
-	debugf("Starting template processing\n")
-	for strings.Contains(content, "{{") {
-		matches := r.FindAllStringSubmatch(content, -1)
-		if len(matches) == 0 {
-			break
-		}
-
-		replaced := false
-		for _, match := range matches {
-			fullMatch := match[0]
-			varName := match[1]
-
-			// Check if this is a plugin call
-			if strings.HasPrefix(varName, "plugin:") {
-				pluginMatches := pluginPattern.FindStringSubmatch(fullMatch)
-				if len(pluginMatches) >= 3 {
-					namespace := pluginMatches[1]
-					operation := pluginMatches[2]
-					value := ""
-					if len(pluginMatches) == 4 {
-						value = pluginMatches[3]
-					}
-
-					debugf("\nPlugin call:\n")
-					debugf("  Namespace: %s\n", namespace)
-					debugf("  Operation: %s\n", operation)
-					debugf("  Value: %s\n", value)
-
-					var result string
-					var err error
-
-					switch namespace {
-					case "text":
-						debugf("Executing text plugin\n")
-						result, err = textPlugin.Apply(operation, value)
-					case "datetime":
-						debugf("Executing datetime plugin\n")
-						result, err = datetimePlugin.Apply(operation, value)
-					case "file":
-						debugf("Executing file plugin\n")
-						result, err = filePlugin.Apply(operation, value)
-						debugf("File plugin result: %#v\n", result)
-					case "fetch":
-						debugf("Executing fetch plugin\n")
-						result, err = fetchPlugin.Apply(operation, value)
-					case "sys":
-						debugf("Executing sys plugin\n")
-						result, err = sysPlugin.Apply(operation, value)
-					default:
-						return "", fmt.Errorf("unknown plugin namespace: %s", namespace)
-					}
-
-					if err != nil {
-						debugf("Plugin error: %v\n", err)
-						return "", fmt.Errorf("plugin %s error: %v", namespace, err)
-					}
-
-					debugf("Plugin result: %s\n", result)
-					content = strings.ReplaceAll(content, fullMatch, result)
-					debugf("Content after replacement: %s\n", content)
-					continue
-				}
-			}
-
-			if pluginMatches := extensionPattern.FindStringSubmatch(fullMatch); len(pluginMatches) >= 3 {
-				name := pluginMatches[1]
-				operation := pluginMatches[2]
-				value := ""
-				if len(pluginMatches) == 4 {
-					value = pluginMatches[3]
-				}
-
-				debugf("\nExtension call:\n")
-				debugf("  Name: %s\n", name)
-				debugf("  Operation: %s\n", operation)
-				debugf("  Value: %s\n", value)
-
-				result, err := extensionManager.ProcessExtension(name, operation, value)
-				if err != nil {
-					return "", fmt.Errorf("extension %s error: %v", name, err)
-				}
-
-				content = strings.ReplaceAll(content, fullMatch, result)
-				replaced = true
-				continue
-			}
-
-			// Handle regular variables and input
-			debugf("Processing variable: %s\n", varName)
-			if varName == "input" {
-				debugf("Replacing {{input}}\n")
-				replaced = true
-				content = strings.ReplaceAll(content, fullMatch, input)
-			} else {
-				if val, ok := variables[varName]; !ok {
-					debugf("Missing variable: %s\n", varName)
-					missingVars = append(missingVars, varName)
-					return "", fmt.Errorf("missing required variable: %s", varName)
-				} else {
-					debugf("Replacing variable %s with value: %s\n", varName, val)
-					content = strings.ReplaceAll(content, fullMatch, val)
-					replaced = true
-				}
-			}
-			if !replaced {
-				return "", fmt.Errorf("template processing stuck - potential infinite loop")
-			}
+// matchTriple extracts the first two required and optional third value from a token
+// pattern of the form {{type:part1:part2(:part3)?}} returning part1, part2, part3 (possibly empty)
+func matchTriple(r *regexp.Regexp, full string) (string, string, string, bool) {
+	parts := r.FindStringSubmatch(full)
+	if len(parts) >= 3 {
+		v := ""
+		if len(parts) == 4 {
+			v = parts[3]
 		}
+		return parts[1], parts[2], v, true
 	}
+	return "", "", "", false
+}

-	debugf("Starting template processing\n")
-	for strings.Contains(content, "{{") {
-		matches := r.FindAllStringSubmatch(content, -1)
+func ApplyTemplate(content string, variables map[string]string, input string) (string, error) {
+	tokenPattern := regexp.MustCompile(`\{\{([^{}]+)\}\}`)
+
+	debugf("Starting template processing with input='%s'\n", input)
+
+	for {
+		if !strings.Contains(content, "{{") {
+			break
+		}
+		matches := tokenPattern.FindAllStringSubmatch(content, -1)
 		if len(matches) == 0 {
 			break
 		}

-		replaced := false
-		for _, match := range matches {
-			fullMatch := match[0]
-			varName := match[1]
+		progress := false
+		for _, m := range matches {
+			full := m[0]
+			raw := m[1]

-			// Check if this is a plugin call
-			if strings.HasPrefix(varName, "plugin:") {
-				pluginMatches := pluginPattern.FindStringSubmatch(fullMatch)
-				if len(pluginMatches) >= 3 {
-					namespace := pluginMatches[1]
-					operation := pluginMatches[2]
-					value := ""
-					if len(pluginMatches) == 4 {
-						value = pluginMatches[3]
+			// Extension call
+			if strings.HasPrefix(raw, "ext:") {
+				if name, operation, value, ok := matchTriple(extensionPattern, full); ok {
+					if strings.Contains(value, InputSentinel) {
+						value = strings.ReplaceAll(value, InputSentinel, input)
+						debugf("Replaced sentinel in extension value with input\n")
 					}
+					debugf("Extension call: name=%s operation=%s value=%s\n", name, operation, value)
+					result, err := extensionManager.ProcessExtension(name, operation, value)
+					if err != nil {
+						return "", fmt.Errorf("extension %s error: %v", name, err)
+					}
+					content = strings.ReplaceAll(content, full, result)
+					progress = true
+					continue
+				}
+			}

-					debugf("\nPlugin call:\n")
-					debugf("  Namespace: %s\n", namespace)
-					debugf("  Operation: %s\n", operation)
-					debugf("  Value: %s\n", value)
-
-					var result string
-					var err error
-
+			// Plugin call
+			if strings.HasPrefix(raw, "plugin:") {
+				if namespace, operation, value, ok := matchTriple(pluginPattern, full); ok {
+					debugf("Plugin call: namespace=%s operation=%s value=%s\n", namespace, operation, value)
+					var (
+						result string
+						err    error
+					)
 					switch namespace {
 					case "text":
 						debugf("Executing text plugin\n")
@@ -203,39 +116,33 @@ func ApplyTemplate(content string, variables map[string]string, input string) (s
 					default:
 						return "", fmt.Errorf("unknown plugin namespace: %s", namespace)
 					}
-
 					if err != nil {
 						debugf("Plugin error: %v\n", err)
 						return "", fmt.Errorf("plugin %s error: %v", namespace, err)
 					}
-
-					debugf("Plugin result: %s\n", result)
-					content = strings.ReplaceAll(content, fullMatch, result)
-					debugf("Content after replacement: %s\n", content)
+					content = strings.ReplaceAll(content, full, result)
+					progress = true
 					continue
 				}
 			}

-			// Handle regular variables and input
-			debugf("Processing variable: %s\n", varName)
-			if varName == "input" {
-				debugf("Replacing {{input}}\n")
-				replaced = true
-				content = strings.ReplaceAll(content, fullMatch, input)
-			} else {
-				if val, ok := variables[varName]; !ok {
-					debugf("Missing variable: %s\n", varName)
-					missingVars = append(missingVars, varName)
-					return "", fmt.Errorf("missing required variable: %s", varName)
-				} else {
-					debugf("Replacing variable %s with value: %s\n", varName, val)
-					content = strings.ReplaceAll(content, fullMatch, val)
-					replaced = true
+			// Variables / input / sentinel
+			switch raw {
+			case "input", InputSentinel:
+				content = strings.ReplaceAll(content, full, input)
+				progress = true
+			default:
+				val, ok := variables[raw]
+				if !ok {
+					return "", fmt.Errorf("missing required variable: %s", raw)
 				}
+				content = strings.ReplaceAll(content, full, val)
+				progress = true
 			}
-			if !replaced {
-				return "", fmt.Errorf("template processing stuck - potential infinite loop")
-			}
+		}
+
+		if !progress {
+			return "", fmt.Errorf("template processing stuck - potential infinite loop")
 		}
 	}

--- a/internal/plugins/template/template_extension_mixed_test.go
+++ b/internal/plugins/template/template_extension_mixed_test.go
@@ -0,0 +1,77 @@
+package template
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+// TestExtensionValueMixedInputAndVariable ensures an extension value mixing {{input}} and another template variable is processed.
+func TestExtensionValueMixedInputAndVariable(t *testing.T) {
+	input := "PRIMARY"
+	variables := map[string]string{
+		"suffix": "SUF",
+	}
+
+	// Build temp extension environment
+	tmp := t.TempDir()
+	configDir := filepath.Join(tmp, ".config", "fabric")
+	extsDir := filepath.Join(configDir, "extensions")
+	binDir := filepath.Join(extsDir, "bin")
+	configsDir := filepath.Join(extsDir, "configs")
+	if err := os.MkdirAll(binDir, 0o755); err != nil {
+		t.Fatalf("mkdir bin: %v", err)
+	}
+	if err := os.MkdirAll(configsDir, 0o755); err != nil {
+		t.Fatalf("mkdir configs: %v", err)
+	}
+
+	scriptPath := filepath.Join(binDir, "mix-echo.sh")
+	// Simple echo script; avoid percent formatting complexities
+	script := "#!/bin/sh\necho VAL=$1\n"
+	if err := os.WriteFile(scriptPath, []byte(script), 0o755); err != nil {
+		t.Fatalf("write script: %v", err)
+	}
+
+	configYAML := "" +
+		"name: mix-echo\n" +
+		"type: executable\n" +
+		"executable: " + scriptPath + "\n" +
+		"description: mixed input/variable test\n" +
+		"version: 1.0.0\n" +
+		"timeout: 5s\n" +
+		"operations:\n" +
+		"  echo:\n" +
+		"    cmd_template: '{{executable}} {{value}}'\n"
+	if err := os.WriteFile(filepath.Join(configsDir, "mix-echo.yaml"), []byte(configYAML), 0o644); err != nil {
+		t.Fatalf("write config: %v", err)
+	}
+
+	// Use a fresh extension manager isolated from global one
+	mgr := NewExtensionManager(configDir)
+	if err := mgr.RegisterExtension(filepath.Join(configsDir, "mix-echo.yaml")); err != nil {
+		// Some environments may not support execution; skip instead of fail hard
+		if strings.Contains(err.Error(), "operation not permitted") {
+			t.Skipf("skipping due to exec restriction: %v", err)
+		}
+		t.Fatalf("register: %v", err)
+	}
+
+	// Temporarily swap global extensionManager for this test
+	prevMgr := extensionManager
+	extensionManager = mgr
+	defer func() { extensionManager = prevMgr }()
+
+	// Template uses input plus a variable inside extension value
+	tmpl := "{{ext:mix-echo:echo:pre-{{input}}-mid-{{suffix}}-post}}"
+
+	out, err := ApplyTemplate(tmpl, variables, input)
+	if err != nil {
+		t.Fatalf("ApplyTemplate error: %v", err)
+	}
+
+	if !strings.Contains(out, "VAL=pre-PRIMARY-mid-SUF-post") {
+		t.Fatalf("unexpected output: %q", out)
+	}
+}
--- a/internal/plugins/template/template_extension_multiple_test.go
+++ b/internal/plugins/template/template_extension_multiple_test.go
@@ -0,0 +1,71 @@
+package template
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+// TestMultipleExtensionsWithInput ensures multiple extension calls each using {{input}} get proper substitution.
+func TestMultipleExtensionsWithInput(t *testing.T) {
+	input := "DATA"
+	variables := map[string]string{}
+
+	tmp := t.TempDir()
+	configDir := filepath.Join(tmp, ".config", "fabric")
+	extsDir := filepath.Join(configDir, "extensions")
+	binDir := filepath.Join(extsDir, "bin")
+	configsDir := filepath.Join(extsDir, "configs")
+	if err := os.MkdirAll(binDir, 0o755); err != nil {
+		t.Fatalf("mkdir bin: %v", err)
+	}
+	if err := os.MkdirAll(configsDir, 0o755); err != nil {
+		t.Fatalf("mkdir configs: %v", err)
+	}
+
+	scriptPath := filepath.Join(binDir, "multi-echo.sh")
+	script := "#!/bin/sh\necho ECHO=$1\n"
+	if err := os.WriteFile(scriptPath, []byte(script), 0o755); err != nil {
+		t.Fatalf("write script: %v", err)
+	}
+
+	configYAML := "" +
+		"name: multi-echo\n" +
+		"type: executable\n" +
+		"executable: " + scriptPath + "\n" +
+		"description: multi echo extension\n" +
+		"version: 1.0.0\n" +
+		"timeout: 5s\n" +
+		"operations:\n" +
+		"  echo:\n" +
+		"    cmd_template: '{{executable}} {{value}}'\n"
+	if err := os.WriteFile(filepath.Join(configsDir, "multi-echo.yaml"), []byte(configYAML), 0o644); err != nil {
+		t.Fatalf("write config: %v", err)
+	}
+
+	mgr := NewExtensionManager(configDir)
+	if err := mgr.RegisterExtension(filepath.Join(configsDir, "multi-echo.yaml")); err != nil {
+		t.Fatalf("register: %v", err)
+	}
+	prev := extensionManager
+	extensionManager = mgr
+	defer func() { extensionManager = prev }()
+
+	tmpl := strings.Join([]string{
+		"First: {{ext:multi-echo:echo:{{input}}}}",
+		"Second: {{ext:multi-echo:echo:{{input}}}}",
+		"Third: {{ext:multi-echo:echo:{{input}}}}",
+	}, " | ")
+
+	out, err := ApplyTemplate(tmpl, variables, input)
+	if err != nil {
+		t.Fatalf("ApplyTemplate error: %v", err)
+	}
+
+	wantCount := 3
+	occ := strings.Count(out, "ECHO=DATA")
+	if occ != wantCount {
+		t.Fatalf("expected %d occurrences of ECHO=DATA, got %d; output=%q", wantCount, occ, out)
+	}
+}
--- a/internal/plugins/template/template_sentinel_test.go
+++ b/internal/plugins/template/template_sentinel_test.go
@@ -0,0 +1,275 @@
+package template
+
+import (
+	"fmt"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+// withTestExtension creates a temporary test extension and runs the test function
+func withTestExtension(t *testing.T, name string, scriptContent string, testFunc func(*ExtensionManager, string)) {
+	t.Helper()
+
+	// Create a temporary directory for test extension
+	tmpDir := t.TempDir()
+	configDir := filepath.Join(tmpDir, ".config", "fabric")
+	extensionsDir := filepath.Join(configDir, "extensions")
+	binDir := filepath.Join(extensionsDir, "bin")
+	configsDir := filepath.Join(extensionsDir, "configs")
+
+	err := os.MkdirAll(binDir, 0755)
+	if err != nil {
+		t.Fatalf("Failed to create bin directory: %v", err)
+	}
+	err = os.MkdirAll(configsDir, 0755)
+	if err != nil {
+		t.Fatalf("Failed to create configs directory: %v", err)
+	}
+
+	// Create a test script
+	scriptPath := filepath.Join(binDir, name+".sh")
+	err = os.WriteFile(scriptPath, []byte(scriptContent), 0755)
+	if err != nil {
+		t.Fatalf("Failed to create test script: %v", err)
+	}
+
+	// Create extension config
+	configPath := filepath.Join(configsDir, name+".yaml")
+	configContent := fmt.Sprintf(`name: %s
+executable: %s
+type: executable
+timeout: "5s"
+description: "Test extension"
+version: "1.0.0"
+
+operations:
+  echo:
+    cmd_template: "{{executable}} {{value}}"
+
+config:
+  output:
+    method: stdout
+`, name, scriptPath)
+	err = os.WriteFile(configPath, []byte(configContent), 0644)
+	if err != nil {
+		t.Fatalf("Failed to create extension config: %v", err)
+	}
+
+	// Initialize extension manager with test config directory
+	mgr := NewExtensionManager(configDir)
+
+	// Register the test extension
+	err = mgr.RegisterExtension(configPath)
+	if err != nil {
+		t.Fatalf("Failed to register extension: %v", err)
+	}
+
+	// Run the test
+	testFunc(mgr, name)
+}
+
+// TestSentinelTokenReplacement tests the fix for the {{input}} sentinel token bug
+// This test verifies that when {{input}} is used inside an extension call,
+// the actual input is passed to the extension, not the sentinel token.
+func TestSentinelTokenReplacement(t *testing.T) {
+	scriptContent := `#!/bin/bash
+echo "RECEIVED: $@"
+`
+
+	withTestExtension(t, "echo-test", scriptContent, func(mgr *ExtensionManager, name string) {
+		// Save and restore global extension manager
+		oldManager := extensionManager
+		defer func() { extensionManager = oldManager }()
+		extensionManager = mgr
+
+		tests := []struct {
+			name           string
+			template       string
+			input          string
+			wantContain    string
+			wantNotContain string
+		}{
+			{
+				name:           "sentinel token with {{input}} in extension value",
+				template:       "{{ext:echo-test:echo:__FABRIC_INPUT_SENTINEL_TOKEN__}}",
+				input:          "test input data",
+				wantContain:    "RECEIVED: test input data",
+				wantNotContain: "__FABRIC_INPUT_SENTINEL_TOKEN__",
+			},
+			{
+				name:           "direct input variable replacement",
+				template:       "{{ext:echo-test:echo:{{input}}}}",
+				input:          "Hello World",
+				wantContain:    "RECEIVED: Hello World",
+				wantNotContain: "{{input}}",
+			},
+			{
+				name:           "sentinel with complex input",
+				template:       "Result: {{ext:echo-test:echo:__FABRIC_INPUT_SENTINEL_TOKEN__}}",
+				input:          "What is AI?",
+				wantContain:    "RECEIVED: What is AI?",
+				wantNotContain: "__FABRIC_INPUT_SENTINEL_TOKEN__",
+			},
+			{
+				name:           "multiple words in input",
+				template:       "{{ext:echo-test:echo:{{input}}}}",
+				input:          "Multiple word input string",
+				wantContain:    "RECEIVED: Multiple word input string",
+				wantNotContain: "{{input}}",
+			},
+		}
+
+		for _, tt := range tests {
+			t.Run(tt.name, func(t *testing.T) {
+				got, err := ApplyTemplate(tt.template, map[string]string{}, tt.input)
+				if err != nil {
+					t.Errorf("ApplyTemplate() error = %v", err)
+					return
+				}
+
+				// Check that result contains expected string
+				if !strings.Contains(got, tt.wantContain) {
+					t.Errorf("ApplyTemplate() = %q, should contain %q", got, tt.wantContain)
+				}
+
+				// Check that result does NOT contain unwanted string
+				if strings.Contains(got, tt.wantNotContain) {
+					t.Errorf("ApplyTemplate() = %q, should NOT contain %q", got, tt.wantNotContain)
+				}
+			})
+		}
+	})
+}
+
+// TestSentinelInVariableProcessing tests that the sentinel token is handled
+// correctly in regular variable processing (not just extensions)
+// Note: The sentinel is only replaced when it appears in extension values,
+// not when used as a standalone variable (which would be a user error)
+func TestSentinelInVariableProcessing(t *testing.T) {
+	tests := []struct {
+		name     string
+		template string
+		vars     map[string]string
+		input    string
+		want     string
+	}{
+		{
+			name:     "input variable works normally",
+			template: "Value: {{input}}",
+			input:    "actual input",
+			want:     "Value: actual input",
+		},
+		{
+			name:     "multiple input references",
+			template: "First: {{input}}, Second: {{input}}",
+			input:    "test",
+			want:     "First: test, Second: test",
+		},
+		{
+			name:     "input with variables",
+			template: "Var: {{name}}, Input: {{input}}",
+			vars:     map[string]string{"name": "TestVar"},
+			input:    "input value",
+			want:     "Var: TestVar, Input: input value",
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := ApplyTemplate(tt.template, tt.vars, tt.input)
+			if err != nil {
+				t.Errorf("ApplyTemplate() error = %v", err)
+				return
+			}
+
+			if got != tt.want {
+				t.Errorf("ApplyTemplate() = %q, want %q", got, tt.want)
+			}
+		})
+	}
+}
+
+// TestExtensionValueWithSentinel specifically tests the extension value
+// sentinel replacement logic
+func TestExtensionValueWithSentinel(t *testing.T) {
+	scriptContent := `#!/bin/bash
+# Output each argument on a separate line
+for arg in "$@"; do
+    echo "ARG: $arg"
+done
+`
+
+	withTestExtension(t, "arg-test", scriptContent, func(mgr *ExtensionManager, name string) {
+		// Save and restore global extension manager
+		oldManager := extensionManager
+		defer func() { extensionManager = oldManager }()
+		extensionManager = mgr
+
+		// Test that sentinel token in extension value gets replaced
+		template := "{{ext:arg-test:echo:prefix-__FABRIC_INPUT_SENTINEL_TOKEN__-suffix}}"
+		input := "MYINPUT"
+
+		got, err := ApplyTemplate(template, map[string]string{}, input)
+		if err != nil {
+			t.Fatalf("ApplyTemplate() error = %v", err)
+		}
+
+		// The sentinel should be replaced with actual input
+		expectedContain := "ARG: prefix-MYINPUT-suffix"
+		if !strings.Contains(got, expectedContain) {
+			t.Errorf("ApplyTemplate() = %q, should contain %q", got, expectedContain)
+		}
+
+		// The sentinel token should NOT appear in output
+		if strings.Contains(got, "__FABRIC_INPUT_SENTINEL_TOKEN__") {
+			t.Errorf("ApplyTemplate() = %q, should NOT contain sentinel token", got)
+		}
+	})
+}
+
+// TestNestedInputInExtension tests the original bug case:
+// {{ext:name:op:{{input}}}} should pass the actual input, not the sentinel
+func TestNestedInputInExtension(t *testing.T) {
+	scriptContent := `#!/bin/bash
+echo "NESTED_TEST: $*"
+`
+
+	withTestExtension(t, "nested-test", scriptContent, func(mgr *ExtensionManager, name string) {
+		// Save and restore global extension manager
+		oldManager := extensionManager
+		defer func() { extensionManager = oldManager }()
+		extensionManager = mgr
+
+		// This is the bug case: {{input}} nested inside extension call
+		// The template processing should:
+		// 1. Replace {{input}} with sentinel during variable protection
+		// 2. Process the extension, replacing sentinel with actual input
+		// 3. Execute extension with actual input, not sentinel
+
+		template := "{{ext:nested-test:echo:{{input}}}}"
+		input := "What is Artificial Intelligence"
+
+		got, err := ApplyTemplate(template, map[string]string{}, input)
+		if err != nil {
+			t.Fatalf("ApplyTemplate() error = %v", err)
+		}
+
+		// Verify the actual input was passed, not the sentinel
+		expectedContain := "NESTED_TEST: What is Artificial Intelligence"
+		if !strings.Contains(got, expectedContain) {
+			t.Errorf("ApplyTemplate() = %q, should contain %q", got, expectedContain)
+		}
+
+		// Verify sentinel token does NOT appear
+		if strings.Contains(got, "__FABRIC_INPUT_SENTINEL_TOKEN__") {
+			t.Errorf("ApplyTemplate() output contains sentinel token (BUG NOT FIXED): %q", got)
+		}
+
+		// Verify {{input}} template tag does NOT appear
+		if strings.Contains(got, "{{input}}") {
+			t.Errorf("ApplyTemplate() output contains unresolved {{input}}: %q", got)
+		}
+	})
+}
--- a/internal/plugins/template/text.go
+++ b/internal/plugins/template/text.go
@@ -16,7 +16,7 @@ func toTitle(s string) string {
 	lower := strings.ToLower(s)
 	runes := []rune(lower)

-	for i := 0; i < len(runes); i++ {
+	for i := range runes {
 		// Capitalize if previous char is non-letter AND
 		// (we're at the end OR next char is not space)
 		if i == 0 || !unicode.IsLetter(runes[i-1]) {
--- a/internal/server/models.go
+++ b/internal/server/models.go
@@ -24,7 +24,7 @@ func (h *ModelsHandler) GetModelNames(c *gin.Context) {
 		return
 	}

-	response := make(map[string]interface{})
+	response := make(map[string]any)
 	vendors := make(map[string][]string)

 	for _, groupItems := range vendorsModels.GroupsItems {
--- a/internal/server/ollama.go
+++ b/internal/server/ollama.go
@@ -102,7 +102,7 @@ func ServeOllama(registry *core.PluginRegistry, address string, version string)
 	// Ollama Endpoints
 	r.GET("/api/tags", typeConversion.ollamaTags)
 	r.GET("/api/version", func(c *gin.Context) {
-		c.Data(200, "application/json", []byte(fmt.Sprintf("{\"%s\"}", version)))
+		c.Data(200, "application/json", fmt.Appendf(nil, "{\"%s\"}", version))
 	})
 	r.POST("/api/chat", typeConversion.ollamaChat)

@@ -224,7 +224,7 @@ func (f APIConvert) ollamaChat(c *gin.Context) {
 		c.JSON(http.StatusInternalServerError, gin.H{"error": "testing endpoint"})
 		return
 	}
-	for _, word := range strings.Split(fabricResponse.Content, " ") {
+	for word := range strings.SplitSeq(fabricResponse.Content, " ") {
 		forwardedResponse = OllamaResponse{
 			Model:     "",
 			CreatedAt: "",
--- a/internal/server/patterns.go
+++ b/internal/server/patterns.go
@@ -1,6 +1,7 @@
 package restapi

 import (
+	"maps"
 	"net/http"

 	"github.com/danielmiessler/fabric/internal/plugins/db/fsdb"
@@ -74,9 +75,7 @@ func (h *PatternsHandler) ApplyPattern(c *gin.Context) {
 			variables[key] = values[0]
 		}
 	}
-	for key, value := range request.Variables {
-		variables[key] = value
-	}
+	maps.Copy(variables, request.Variables)

 	pattern, err := h.patterns.GetApplyVariables(name, variables, request.Input)
 	if err != nil {
--- a/internal/tools/youtube/youtube.go
+++ b/internal/tools/youtube/youtube.go
@@ -26,10 +26,11 @@ import (
 	"strings"
 	"time"

+	"github.com/danielmiessler/fabric/internal/i18n"
+	debuglog "github.com/danielmiessler/fabric/internal/log"
 	"github.com/danielmiessler/fabric/internal/plugins"
 	"github.com/kballard/go-shellquote"

-	debuglog "github.com/danielmiessler/fabric/internal/log"
 	"google.golang.org/api/option"
 	"google.golang.org/api/youtube/v3"
 )
@@ -69,7 +70,7 @@ func NewYouTube() (ret *YouTube) {
 		EnvNamePrefix:    plugins.BuildEnvVariablePrefix(label),
 	}

-	ret.ApiKey = ret.AddSetupQuestion("API key", true)
+	ret.ApiKey = ret.AddSetupQuestion("API key", false)

 	return
 }
@@ -85,7 +86,7 @@ type YouTube struct {
 func (o *YouTube) initService() (err error) {
 	if o.service == nil {
 		if o.ApiKey.Value == "" {
-			err = fmt.Errorf("YouTube API key required for comments and metadata. Run 'fabric --setup' to configure")
+			err = fmt.Errorf("%s", i18n.T("youtube_api_key_required"))
 			return
 		}
 		o.normalizeRegex = regexp.MustCompile(`[^a-zA-Z0-9]+`)
@@ -109,41 +110,66 @@ func (o *YouTube) GetVideoOrPlaylistId(url string) (videoId string, playlistId s
 	}

 	if videoId == "" && playlistId == "" {
-		err = fmt.Errorf("invalid YouTube URL, can't get video or playlist ID: '%s'", url)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_invalid_url"), url))
 	}
 	return
 }

-func (o *YouTube) GrabTranscriptForUrl(url string, language string) (ret string, err error) {
-	var videoId string
+// extractAndValidateVideoId extracts a video ID from the given URL and validates
+// that the URL points to a video rather than a playlist-only resource.
+// It returns an error if the URL is invalid or contains only playlist information.
+func (o *YouTube) extractAndValidateVideoId(url string) (videoId string, err error) {
 	var playlistId string
 	if videoId, playlistId, err = o.GetVideoOrPlaylistId(url); err != nil {
-		return
-	} else if videoId == "" && playlistId != "" {
-		err = fmt.Errorf("URL is a playlist, not a video")
+		return "", err
+	}
+	if videoId == "" && playlistId != "" {
+		return "", fmt.Errorf("%s", i18n.T("youtube_url_is_playlist_not_video"))
+	}
+	if videoId == "" {
+		return "", fmt.Errorf("%s", i18n.T("youtube_no_video_id_found"))
+	}
+	return videoId, nil
+}
+
+func (o *YouTube) GrabTranscriptForUrl(url string, language string) (ret string, err error) {
+	var videoId string
+	if videoId, err = o.extractAndValidateVideoId(url); err != nil {
 		return
 	}
-
 	return o.GrabTranscript(videoId, language)
 }

+// GrabTranscript retrieves the transcript for the specified video ID using yt-dlp.
+// The language parameter specifies the preferred subtitle language code (e.g., "en", "es").
+// It returns the transcript text or an error if the transcript cannot be retrieved.
 func (o *YouTube) GrabTranscript(videoId string, language string) (ret string, err error) {
-	// Use yt-dlp for reliable transcript extraction
 	return o.GrabTranscriptWithArgs(videoId, language, "")
 }

+// GrabTranscriptWithArgs retrieves the transcript for the specified video ID using yt-dlp
+// with custom command-line arguments. The language parameter specifies the preferred subtitle
+// language code. The additionalArgs parameter allows passing extra yt-dlp options like
+// "--cookies-from-browser brave" for authentication.
+// It returns the transcript text or an error if the transcript cannot be retrieved.
 func (o *YouTube) GrabTranscriptWithArgs(videoId string, language string, additionalArgs string) (ret string, err error) {
-	// Use yt-dlp for reliable transcript extraction
 	return o.tryMethodYtDlp(videoId, language, additionalArgs)
 }

+// GrabTranscriptWithTimestamps retrieves the transcript with timestamps for the specified
+// video ID using yt-dlp. The language parameter specifies the preferred subtitle language code.
+// Each line in the returned transcript is prefixed with a timestamp in [HH:MM:SS] format.
+// It returns the timestamped transcript text or an error if the transcript cannot be retrieved.
 func (o *YouTube) GrabTranscriptWithTimestamps(videoId string, language string) (ret string, err error) {
-	// Use yt-dlp for reliable transcript extraction with timestamps
 	return o.GrabTranscriptWithTimestampsWithArgs(videoId, language, "")
 }

+// GrabTranscriptWithTimestampsWithArgs retrieves the transcript with timestamps for the specified
+// video ID using yt-dlp with custom command-line arguments. The language parameter specifies the
+// preferred subtitle language code. The additionalArgs parameter allows passing extra yt-dlp options.
+// Each line in the returned transcript is prefixed with a timestamp in [HH:MM:SS] format.
+// It returns the timestamped transcript text or an error if the transcript cannot be retrieved.
 func (o *YouTube) GrabTranscriptWithTimestampsWithArgs(videoId string, language string, additionalArgs string) (ret string, err error) {
-	// Use yt-dlp for reliable transcript extraction with timestamps
 	return o.tryMethodYtDlpWithTimestamps(videoId, language, additionalArgs)
 }

@@ -153,10 +179,10 @@ func detectError(ytOutput io.Reader) error {
 		curLine := scanner.Text()
 		debuglog.Debug(debuglog.Trace, "%s\n", curLine)
 		errorMessages := map[string]string{
-			"429":                                 "YouTube rate limit exceeded. Try again later or use different yt-dlp arguments like '--sleep-requests 1' to slow down requests.",
-			"Too Many Requests":                   "YouTube rate limit exceeded. Try again later or use different yt-dlp arguments like '--sleep-requests 1' to slow down requests.",
-			"Sign in to confirm you're not a bot": "YouTube requires authentication (bot detection). Use --yt-dlp-args='--cookies-from-browser BROWSER' where BROWSER is chrome, firefox, brave, etc.",
-			"Use --cookies-from-browser":          "YouTube requires authentication (bot detection). Use --yt-dlp-args='--cookies-from-browser BROWSER' where BROWSER is chrome, firefox, brave, etc.",
+			"429":                                 i18n.T("youtube_rate_limit_exceeded"),
+			"Too Many Requests":                   i18n.T("youtube_rate_limit_exceeded"),
+			"Sign in to confirm you're not a bot": i18n.T("youtube_auth_required_bot_detection"),
+			"Use --cookies-from-browser":          i18n.T("youtube_auth_required_bot_detection"),
 		}

 		for key, message := range errorMessages {
@@ -166,7 +192,7 @@ func detectError(ytOutput io.Reader) error {
 		}
 	}
 	if err := scanner.Err(); err != nil {
-		return fmt.Errorf("Error reading yt-dlp stderr")
+		return fmt.Errorf("%s", i18n.T("youtube_ytdlp_stderr_error"))
 	}
 	return nil
 }
@@ -192,14 +218,14 @@ func noLangs(args []string) []string {
 func (o *YouTube) tryMethodYtDlpInternal(videoId string, language string, additionalArgs string, processVTTFileFunc func(filename string) (string, error)) (ret string, err error) {
 	// Check if yt-dlp is available
 	if _, err = exec.LookPath("yt-dlp"); err != nil {
-		err = fmt.Errorf("yt-dlp not found in PATH. Please install yt-dlp to use YouTube transcript functionality")
+		err = fmt.Errorf("%s", i18n.T("youtube_ytdlp_not_found"))
 		return
 	}

 	// Create a temporary directory for yt-dlp output (cross-platform)
 	tempDir := filepath.Join(os.TempDir(), "fabric-youtube-"+videoId)
 	if err = os.MkdirAll(tempDir, 0755); err != nil {
-		err = fmt.Errorf("failed to create temp directory: %v", err)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_failed_create_temp_dir"), err))
 		return
 	}
 	defer os.RemoveAll(tempDir)
@@ -231,7 +257,7 @@ func (o *YouTube) tryMethodYtDlpInternal(videoId string, language string, additi
 	if additionalArgs != "" {
 		additionalArgsList, err := shellquote.Split(additionalArgs)
 		if err != nil {
-			return "", fmt.Errorf("invalid yt-dlp arguments: %v", err)
+			return "", fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_invalid_ytdlp_arguments"), err))
 		}
 		args = append(args, additionalArgsList...)
 	}
@@ -302,7 +328,7 @@ func (o *YouTube) readAndCleanVTTFile(filename string) (ret string, err error) {

 	ret = strings.TrimSpace(textBuilder.String())
 	if ret == "" {
-		err = fmt.Errorf("no transcript content found in VTT file")
+		err = fmt.Errorf("%s", i18n.T("youtube_no_transcript_content"))
 	}
 	return
 }
@@ -372,7 +398,7 @@ func (o *YouTube) readAndFormatVTTWithTimestamps(filename string) (ret string, e

 	ret = strings.TrimSpace(textBuilder.String())
 	if ret == "" {
-		err = fmt.Errorf("no transcript content found in VTT file")
+		err = fmt.Errorf("%s", i18n.T("youtube_no_transcript_content"))
 	}
 	return
 }
@@ -418,7 +444,7 @@ func shouldIncludeRepeat(lastTimestamp, currentTimestamp string) bool {
 func parseTimestampToSeconds(timestamp string) (int, error) {
 	parts := strings.Split(timestamp, ":")
 	if len(parts) < 2 || len(parts) > 3 {
-		return 0, fmt.Errorf("invalid timestamp format: %s", timestamp)
+		return 0, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_invalid_timestamp_format"), timestamp))
 	}

 	var hours, minutes, seconds int
@@ -448,20 +474,27 @@ func parseTimestampToSeconds(timestamp string) (int, error) {
 	return hours*3600 + minutes*60 + seconds, nil
 }

-func parseSeconds(seconds_str string) (int, error) {
-	var seconds int
-	var err error
-	if strings.Contains(seconds_str, ".") {
-		// Handle fractional seconds
-		second_parts := strings.Split(seconds_str, ".")
-		if seconds, err = strconv.Atoi(second_parts[0]); err != nil {
-			return 0, err
-		}
-	} else {
-		if seconds, err = strconv.Atoi(seconds_str); err != nil {
-			return 0, err
+func parseSeconds(secondsStr string) (int, error) {
+	if secondsStr == "" {
+		return 0, fmt.Errorf("%s", i18n.T("youtube_empty_seconds_string"))
+	}
+
+	// Extract integer part (before decimal point if present)
+	intPart := secondsStr
+	if idx := strings.Index(secondsStr, "."); idx != -1 {
+		if idx == 0 {
+			// Handle cases like ".5" -> treat as "0"
+			intPart = "0"
+		} else {
+			intPart = secondsStr[:idx]
 		}
 	}
+
+	seconds, err := strconv.Atoi(intPart)
+	if err != nil {
+		return 0, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_invalid_seconds_format"), secondsStr, err))
+	}
+
 	return seconds, nil
 }

@@ -497,11 +530,7 @@ func (o *YouTube) GrabDurationForUrl(url string) (ret int, err error) {
 	}

 	var videoId string
-	var playlistId string
-	if videoId, playlistId, err = o.GetVideoOrPlaylistId(url); err != nil {
-		return
-	} else if videoId == "" && playlistId != "" {
-		err = fmt.Errorf("URL is a playlist, not a video")
+	if videoId, err = o.extractAndValidateVideoId(url); err != nil {
 		return
 	}
 	return o.GrabDuration(videoId)
@@ -510,7 +539,7 @@ func (o *YouTube) GrabDurationForUrl(url string) (ret int, err error) {
 func (o *YouTube) GrabDuration(videoId string) (ret int, err error) {
 	var videoResponse *youtube.VideoListResponse
 	if videoResponse, err = o.service.Videos.List([]string{"contentDetails"}).Id(videoId).Do(); err != nil {
-		err = fmt.Errorf("error getting video details: %v", err)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_getting_video_details"), err))
 		return
 	}

@@ -518,7 +547,7 @@ func (o *YouTube) GrabDuration(videoId string) (ret int, err error) {

 	matches := durationRegex.FindStringSubmatch(durationStr)
 	if len(matches) == 0 {
-		return 0, fmt.Errorf("invalid duration string: %s", durationStr)
+		return 0, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_invalid_duration_string"), durationStr))
 	}

 	hours, _ := strconv.Atoi(matches[1])
@@ -532,11 +561,7 @@ func (o *YouTube) GrabDuration(videoId string) (ret int, err error) {

 func (o *YouTube) Grab(url string, options *Options) (ret *VideoInfo, err error) {
 	var videoId string
-	var playlistId string
-	if videoId, playlistId, err = o.GetVideoOrPlaylistId(url); err != nil {
-		return
-	} else if videoId == "" && playlistId != "" {
-		err = fmt.Errorf("URL is a playlist, not a video")
+	if videoId, err = o.extractAndValidateVideoId(url); err != nil {
 		return
 	}

@@ -544,14 +569,14 @@ func (o *YouTube) Grab(url string, options *Options) (ret *VideoInfo, err error)

 	if options.Metadata {
 		if ret.Metadata, err = o.GrabMetadata(videoId); err != nil {
-			err = fmt.Errorf("error getting video metadata: %v", err)
+			err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_getting_metadata"), err))
 			return
 		}
 	}

 	if options.Duration {
 		if ret.Duration, err = o.GrabDuration(videoId); err != nil {
-			err = fmt.Errorf("error parsing video duration: %v", err)
+			err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_parsing_duration"), err))
 			return
 		}

@@ -559,7 +584,7 @@ func (o *YouTube) Grab(url string, options *Options) (ret *VideoInfo, err error)

 	if options.Comments {
 		if ret.Comments, err = o.GrabComments(videoId); err != nil {
-			err = fmt.Errorf("error getting comments: %v", err)
+			err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_getting_comments"), err))
 			return
 		}
 	}
@@ -643,12 +668,12 @@ func (o *YouTube) SaveVideosToCSV(filename string, videos []*VideoMeta) (err err
 func (o *YouTube) FetchAndSavePlaylist(playlistID, filename string) (err error) {
 	var videos []*VideoMeta
 	if videos, err = o.FetchPlaylistVideos(playlistID); err != nil {
-		err = fmt.Errorf("error fetching playlist videos: %v", err)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("error_fetching_playlist_videos"), err))
 		return
 	}

 	if err = o.SaveVideosToCSV(filename, videos); err != nil {
-		err = fmt.Errorf("error saving videos to CSV: %v", err)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_saving_csv"), err))
 		return
 	}

@@ -659,7 +684,7 @@ func (o *YouTube) FetchAndSavePlaylist(playlistID, filename string) (err error)
 func (o *YouTube) FetchAndPrintPlaylist(playlistID string) (err error) {
 	var videos []*VideoMeta
 	if videos, err = o.FetchPlaylistVideos(playlistID); err != nil {
-		err = fmt.Errorf("error fetching playlist videos: %v", err)
+		err = fmt.Errorf("%s", fmt.Sprintf(i18n.T("error_fetching_playlist_videos"), err))
 		return
 	}

@@ -694,11 +719,11 @@ func (o *YouTube) findVTTFilesWithFallback(dir, requestedLanguage string) ([]str
 	})

 	if err != nil {
-		return nil, fmt.Errorf("failed to walk directory: %v", err)
+		return nil, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_failed_walk_directory"), err))
 	}

 	if len(vttFiles) == 0 {
-		return nil, fmt.Errorf("no VTT files found in directory")
+		return nil, fmt.Errorf("%s", i18n.T("youtube_no_vtt_files_found"))
 	}

 	// If no specific language requested, return the first file
@@ -769,11 +794,11 @@ func (o *YouTube) GrabMetadata(videoId string) (metadata *VideoMetadata, err err
 	call := o.service.Videos.List([]string{"snippet", "statistics"}).Id(videoId)
 	var response *youtube.VideoListResponse
 	if response, err = call.Do(); err != nil {
-		return nil, fmt.Errorf("error getting video metadata: %v", err)
+		return nil, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_error_getting_metadata"), err))
 	}

 	if len(response.Items) == 0 {
-		return nil, fmt.Errorf("no video found with ID: %s", videoId)
+		return nil, fmt.Errorf("%s", fmt.Sprintf(i18n.T("youtube_no_video_found_with_id"), videoId))
 	}

 	video := response.Items[0]
--- a/internal/tools/youtube/youtube_optional_test.go
+++ b/internal/tools/youtube/youtube_optional_test.go
@@ -0,0 +1,19 @@
+package youtube
+
+import "testing"
+
+func TestNewYouTubeApiKeyOptional(t *testing.T) {
+	yt := NewYouTube()
+
+	if yt.ApiKey == nil {
+		t.Fatal("expected API key setup question to be initialized")
+	}
+
+	if yt.ApiKey.Required {
+		t.Fatalf("expected YouTube API key to be optional, but it is marked as required")
+	}
+
+	if !yt.IsConfigured() {
+		t.Fatalf("expected YouTube plugin to be considered configured without an API key")
+	}
+}
--- a/internal/tools/youtube/youtube_test.go
+++ b/internal/tools/youtube/youtube_test.go
@@ -0,0 +1,168 @@
+package youtube
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestParseSeconds(t *testing.T) {
+	tests := []struct {
+		name    string
+		input   string
+		want    int
+		wantErr bool
+	}{
+		{
+			name:    "integer seconds",
+			input:   "42",
+			want:    42,
+			wantErr: false,
+		},
+		{
+			name:    "fractional seconds",
+			input:   "42.567",
+			want:    42,
+			wantErr: false,
+		},
+		{
+			name:    "zero",
+			input:   "0",
+			want:    0,
+			wantErr: false,
+		},
+		{
+			name:    "zero with fraction",
+			input:   "0.999",
+			want:    0,
+			wantErr: false,
+		},
+		{
+			name:    "decimal point at start",
+			input:   ".5",
+			want:    0,
+			wantErr: false,
+		},
+		{
+			name:    "invalid input",
+			input:   "abc",
+			want:    0,
+			wantErr: true,
+		},
+		{
+			name:    "empty string",
+			input:   "",
+			want:    0,
+			wantErr: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := parseSeconds(tt.input)
+
+			// Check error condition
+			if tt.wantErr {
+				if err == nil {
+					t.Errorf("parseSeconds(%q) expected error but got none", tt.input)
+				}
+				return
+			}
+
+			// Check success condition
+			if err != nil {
+				t.Fatalf("parseSeconds(%q) unexpected error: %v", tt.input, err)
+			}
+
+			if got != tt.want {
+				t.Errorf("parseSeconds(%q) = %d, want %d", tt.input, got, tt.want)
+			}
+		})
+	}
+}
+
+func TestExtractAndValidateVideoId(t *testing.T) {
+	yt := NewYouTube()
+
+	tests := []struct {
+		name      string
+		url       string
+		wantId    string
+		wantError bool
+		errorMsg  string
+	}{
+		{
+			name:      "valid video URL",
+			url:       "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
+			wantId:    "dQw4w9WgXcQ",
+			wantError: false,
+		},
+		{
+			name:      "valid short URL",
+			url:       "https://youtu.be/dQw4w9WgXcQ",
+			wantId:    "dQw4w9WgXcQ",
+			wantError: false,
+		},
+		{
+			name:      "video with playlist URL - should extract video",
+			url:       "https://www.youtube.com/watch?v=dQw4w9WgXcQ&list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf",
+			wantId:    "dQw4w9WgXcQ",
+			wantError: false,
+		},
+		{
+			name:      "playlist-only URL",
+			url:       "https://www.youtube.com/playlist?list=PLrAXtmErZgOeiKm4sgNOknGvNjby9efdf",
+			wantId:    "",
+			wantError: true,
+			errorMsg:  "URL is a playlist, not a video",
+		},
+		{
+			name:      "invalid URL",
+			url:       "https://example.com",
+			wantId:    "",
+			wantError: true,
+			errorMsg:  "invalid YouTube URL",
+		},
+		{
+			name:      "empty URL",
+			url:       "",
+			wantId:    "",
+			wantError: true,
+		},
+		{
+			name:      "malformed URL",
+			url:       "not-a-url",
+			wantId:    "",
+			wantError: true,
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := yt.extractAndValidateVideoId(tt.url)
+
+			if tt.wantError {
+				if err == nil {
+					t.Errorf("extractAndValidateVideoId(%q) expected error but got none", tt.url)
+					return
+				}
+				if tt.errorMsg != "" && !strings.Contains(err.Error(), tt.errorMsg) {
+					t.Errorf("extractAndValidateVideoId(%q) error = %v, want error containing %q", tt.url, err, tt.errorMsg)
+				}
+				// Verify empty videoId is returned on error
+				if got != "" {
+					t.Errorf("extractAndValidateVideoId(%q) returned videoId %q on error, want empty string", tt.url, got)
+				}
+				return
+			}
+
+			if err != nil {
+				t.Errorf("extractAndValidateVideoId(%q) unexpected error = %v", tt.url, err)
+				return
+			}
+
+			if got != tt.wantId {
+				t.Errorf("extractAndValidateVideoId(%q) = %q, want %q", tt.url, got, tt.wantId)
+			}
+		})
+	}
+}
--- a/internal/util/groups_items.go
+++ b/internal/util/groups_items.go
@@ -133,7 +133,7 @@ func (o *GroupsItemsSelector[I]) Print(shellCompleteList bool) {

 func (o *GroupsItemsSelector[I]) HasGroup(group string) (ret bool) {
 	for _, groupItems := range o.GroupsItems {
-		if ret = groupItems.Group == group; ret {
+		if ret = strings.EqualFold(groupItems.Group, group); ret {
 			break
 		}
 	}
@@ -146,7 +146,7 @@ func (o *GroupsItemsSelector[I]) FindGroupsByItemFirst(item I) (ret string) {
 	for _, groupItems := range o.GroupsItems {
 		if groupItems.ContainsItemBy(func(groupItem I) bool {
 			groupItemKey := o.GetItemKey(groupItem)
-			return groupItemKey == itemKey
+			return strings.EqualFold(groupItemKey, itemKey)
 		}) {
 			ret = groupItems.Group
 			break
@@ -161,7 +161,7 @@ func (o *GroupsItemsSelector[I]) FindGroupsByItem(item I) (groups []string) {
 	for _, groupItems := range o.GroupsItems {
 		if groupItems.ContainsItemBy(func(groupItem I) bool {
 			groupItemKey := o.GetItemKey(groupItem)
-			return groupItemKey == itemKey
+			return strings.EqualFold(groupItemKey, itemKey)
 		}) {
 			groups = append(groups, groupItems.Group)
 		}
--- a/nix/pkgs/fabric/gomod2nix.toml
+++ b/nix/pkgs/fabric/gomod2nix.toml
@@ -32,8 +32,8 @@ schema = 3
    version = "v1.3.3"
    hash = "sha256-jv7ZshpSd7FZzKKN6hqlUgiR8C3y85zNIS/hq7g76Ho="
  [mod."github.com/anthropics/anthropic-sdk-go"]
-    version = "v1.13.0"
-    hash = "sha256-Mbi37aubaObndySkfO3tE6c8bUJaJugG1E8IM4AxW84="
+    version = "v1.19.0"
+    hash = "sha256-ubYeau5XL0tx4c/79L58rzJGOdOWs9z6WQOtN6mpgxw="
  [mod."github.com/araddon/dateparse"]
    version = "v0.0.0-20210429162001-6b43995a97de"
    hash = "sha256-UuX84naeRGMsFOgIgRoBHG5sNy1CzBkWPKmd6VbLwFw="
@@ -236,8 +236,8 @@ schema = 3
    version = "v0.11.7"
    hash = "sha256-3Wn1JWmil0aQQ2I/r398HbnUsi8ADoroqNyPziuxn/c="
  [mod."github.com/openai/openai-go"]
-    version = "v1.8.2"
-    hash = "sha256-O8aV3zEj6o8kIlzlkYaTW4RzvwR3qNUBYiN8SuTM1R0="
+    version = "v1.12.0"
+    hash = "sha256-JHLlKvDwERPf728GUXBsKU58ODgCxcxEe9TKJTGAG1w="
  [mod."github.com/otiai10/copy"]
    version = "v1.14.1"
    hash = "sha256-8RR7u17SbYg9AeBXVHIv5ZMU+kHmOcx0rLUKyz6YtU0="
@@ -317,26 +317,26 @@ schema = 3
    version = "v0.18.0"
    hash = "sha256-tUpUPERjmRi7zldj0oPlnbnBhEkcI9iQGvP1HqlsK10="
  [mod."golang.org/x/crypto"]
-    version = "v0.41.0"
-    hash = "sha256-o5Di0lsFmYnXl7a5MBTqmN9vXMCRpE9ay71C1Ar8jEY="
+    version = "v0.45.0"
+    hash = "sha256-IpNesJYxFcs2jGvagwJrUD/gsJfA3UiETjQwYByXxSY="
  [mod."golang.org/x/exp"]
    version = "v0.0.0-20250531010427-b6e5de432a8b"
    hash = "sha256-QaFfjyB+pogCkUkJskR9xnXwkCOU828XJRrzwwLm6Ms="
  [mod."golang.org/x/net"]
-    version = "v0.43.0"
-    hash = "sha256-bf3iQFrsC8BoarVaS0uSspEFAcr1zHp1uziTtBpwV34="
+    version = "v0.47.0"
+    hash = "sha256-2qFgCd0YfNCGkLrf+xvnhQtKjSe8CymMdLlN3svUYTg="
  [mod."golang.org/x/oauth2"]
    version = "v0.30.0"
    hash = "sha256-btD7BUtQpOswusZY5qIU90uDo38buVrQ0tmmQ8qNHDg="
  [mod."golang.org/x/sync"]
-    version = "v0.16.0"
-    hash = "sha256-sqKDRESeMzLe0jWGWltLZL/JIgrn0XaIeBWCzVN3Bks="
+    version = "v0.18.0"
+    hash = "sha256-S8o6y7GOaYWeq+TzT8BB6T+1mg82Mu08V0TL3ukJprg="
  [mod."golang.org/x/sys"]
-    version = "v0.35.0"
-    hash = "sha256-ZKM8pesQE6NAFZeKQ84oPn5JMhGr8g4TSwLYAsHMGSI="
+    version = "v0.38.0"
+    hash = "sha256-1+i5EaG3JwH3KMtefzJLG5R6jbOeJM4GK3/LHBVnSy0="
  [mod."golang.org/x/text"]
-    version = "v0.28.0"
-    hash = "sha256-8UlJniGK+km4Hmrw6XMxELnExgrih7+z8tU26Cntmto="
+    version = "v0.31.0"
+    hash = "sha256-AT46RrSmV6+/d5FDhs9fPwYzmQ7WSo+YL9tPfhREwLw="
  [mod."google.golang.org/api"]
    version = "v0.247.0"
    hash = "sha256-UzTtydHmNqh1OXbxcN5qNKQxb5dV6h2Mo6DH4P219Ec="
--- a/nix/pkgs/fabric/version.nix
+++ b/nix/pkgs/fabric/version.nix
@@ -1 +1 @@
-"1.4.320"
+"1.4.349"
--- a/scripts/pattern_descriptions/extract_patterns.py
+++ b/scripts/pattern_descriptions/extract_patterns.py
@@ -3,6 +3,7 @@
 """Extracts pattern information from the ~/.config/fabric/patterns directory,
 creates JSON files for pattern extracts and descriptions, and updates web static files.
 """
+
 import os
 import json
 import shutil
@@ -33,7 +34,13 @@ def get_pattern_extract(pattern_path):
 def extract_pattern_info():
    """Extract pattern information from the patterns directory"""
    script_dir = os.path.dirname(os.path.abspath(__file__))
-    patterns_dir = os.path.expanduser("~/.config/fabric/patterns")
+    local_patterns_dir = os.path.join(script_dir, "..", "..", "data", "patterns")
+
+    if os.path.exists(local_patterns_dir):
+        patterns_dir = local_patterns_dir
+    else:
+        patterns_dir = os.path.expanduser("~/.config/fabric/patterns")
+
    print(f"\nScanning patterns directory: {patterns_dir}")

    extracts_path = os.path.join(script_dir, "pattern_extracts.json")
--- a/scripts/pattern_descriptions/pattern_descriptions.json
+++ b/scripts/pattern_descriptions/pattern_descriptions.json
@@ -159,7 +159,8 @@
      "tags": [
        "ANALYSIS",
        "STRATEGY",
-        "SELF"
+        "SELF",
+        "WELLNESS"
      ]
    },
    {
@@ -744,7 +745,8 @@
      "tags": [
        "ANALYSIS",
        "RESEARCH",
-        "SELF"
+        "SELF",
+        "WELLNESS"
      ]
    },
    {
@@ -1060,7 +1062,8 @@
      "tags": [
        "EXTRACT",
        "SELF",
-        "WISDOM"
+        "WISDOM",
+        "WELLNESS"
      ]
    },
    {
@@ -1098,14 +1101,6 @@
        "REVIEW"
      ]
    },
-    {
-      "patternName": "get_youtube_rss",
-      "description": "Generate RSS feed URLs for YouTube channels.",
-      "tags": [
-        "CONVERSION",
-        "DEVELOPMENT"
-      ]
-    },
    {
      "patternName": "humanize",
      "description": "Transform technical content into approachable language.",
@@ -1235,7 +1230,8 @@
      "tags": [
        "ANALYSIS",
        "LEARNING",
-        "SELF"
+        "SELF",
+        "WELLNESS"
      ]
    },
    {
@@ -1544,7 +1540,8 @@
      "description": "Generate personalized messages of encouragement.",
      "tags": [
        "WRITING",
-        "SELF"
+        "SELF",
+        "WELLNESS"
      ]
    },
    {
@@ -1868,7 +1865,8 @@
      "description": "Analyze a psychological profile, pinpoint issues and strengths, and deliver compassionate, structured strategies for spiritual, mental, and life improvement.",
      "tags": [
        "ANALYSIS",
-        "SELF"
+        "SELF",
+        "WELLNESS"
      ]
    },
    {
@@ -1886,6 +1884,54 @@
        "ANALYSIS",
        "WRITING"
      ]
+    },
+    {
+      "patternName": "fix_typos",
+      "description": "Proofreads and corrects typos, spelling, grammar, and punctuation errors.",
+      "tags": [
+        "WRITING"
+      ]
+    },
+    {
+      "patternName": "model_as_sherlock_freud",
+      "description": "Builds psychological models using detective reasoning and psychoanalytic insight.",
+      "tags": [
+        "ANALYSIS",
+        "SELF",
+        "WELLNESS"
+      ]
+    },
+    {
+      "patternName": "predict_person_actions",
+      "description": "Predicts behavioral responses based on psychological profiles and challenges",
+      "tags": [
+        "ANALYSIS",
+        "SELF",
+        "WELLNESS"
+      ]
+    },
+    {
+      "patternName": "recommend_yoga_practice",
+      "description": "Provides personalized yoga sequences, meditation guidance, and holistic lifestyle advice based on individual profiles.",
+      "tags": [
+        "WELLNESS",
+        "SELF"
+      ]
+    },
+    {
+      "patternName": "create_conceptmap",
+      "description": "Transforms unstructured text or markdown content into an interactive HTML concept map using Vis.js by extracting key concepts and their logical relationships.",
+      "tags": [
+        "VISUALIZE"
+      ]
+    },
+    {
+      "patternName": "concall_summary",
+      "description": "Extract strategic insights from earnings transcripts for investors.",
+      "tags": [
+        "SUMMARIZE",
+        "BUSINESS"
+      ]
    }
  ]
 }
--- a/scripts/pattern_descriptions/pattern_extracts.json
+++ b/scripts/pattern_descriptions/pattern_extracts.json
@@ -540,10 +540,6 @@
      "patternName": "get_wow_per_minute",
      "pattern_extract": "# IDENTITY\n\nYou are an expert at determining the wow-factor of content as measured per minute of content, as determined by the steps below.\n\n# GOALS\n\n- The goal is to determine how densely packed the content is with wow-factor. Note that wow-factor can come from multiple types of wow, such as surprise, novelty, insight, value, and wisdom, and also from multiple types of content such as business, science, art, or philosophy.\n\n- The goal is to determine how rewarding this content will be for a viewer in terms of how often they'll be surprised, learn something new, gain insight, find practical value, or gain wisdom.\n\n# STEPS\n\n- Fully and deeply consume the content at least 319 times, using different interpretive perspectives each time.\n\n- Construct a giant virtual whiteboard in your mind.\n\n- Extract the ideas being presented in the content and place them on your giant virtual whiteboard.\n\n- Extract the novelty of those ideas and place them on your giant virtual whiteboard.\n\n- Extract the insights from those ideas and place them on your giant virtual whiteboard.\n\n- Extract the value of those ideas and place them on your giant virtual whiteboard.\n\n- Extract the wisdom of those ideas and place them on your giant virtual whiteboard."
    },
-    {
-      "patternName": "get_youtube_rss",
-      "pattern_extract": "# IDENTITY AND GOALS\n\nYou are a YouTube infrastructure expert that returns YouTube channel RSS URLs.\n\nYou take any input in, especially YouTube channel IDs, or full URLs, and return the RSS URL for that channel.\n\n# STEPS\n\nHere is the structure for YouTube RSS URLs and their relation to the channel ID and or channel URL:\n\nIf the channel URL is https://www.youtube.com/channel/UCnCikd0s4i9KoDtaHPlK-JA, the RSS URL is https://www.youtube.com/feeds/videos.xml?channel_id=UCnCikd0s4i9KoDtaHPlK-JA\n\n- Extract the channel ID from the channel URL.\n\n- Construct the RSS URL using the channel ID.\n\n- Output the RSS URL.\n\n# OUTPUT\n\n- Output only the RSS URL and nothing else.\n\n- Don't complain, just do it.\n\n# INPUT"
-    },
    {
      "patternName": "humanize",
      "pattern_extract": "# IDENTITY and PURPOSE\n\nYou are a real person whose job is to make text sound natural, conversational, and relatable, just like how an average person talks or writes. Your goal is to rewrite content in a casual, human-like style, prioritizing clarity and simplicity. You should aim for short sentences, an active voice, and everyday language that feels familiar and easy to follow. Avoid long, complex sentences or technical jargon. Instead, focus on breaking ideas into smaller, easy-to-understand parts. Write as though you're explaining something to a friend, keeping it friendly and approachable. Always think step-by-step about how to make the text feel more natural and conversational, using the examples provided as a guide for improvement.\n\nWhile rewriting, ensure the original meaning and tone are preserved. Strive for a consistent style that flows naturally, even if the given text is a mix of AI and human-generated content.\n\n# YOUR TASK\n\nYour task is to rewrite the given AI-generated text to make it sound like it was written by a real person. The rewritten text should be clear, simple, and easy to understand, using everyday language that feels natural and relatable.\n\n- Focus on clarity: Make sure the text is straightforward and avoids unnecessary complexity.\n- Keep it simple: Use common words and phrases that anyone can understand.\n- Prioritize short sentences: Break down long, complicated sentences into smaller, more digestible ones.\n- Maintain context: Ensure that the rewritten text accurately reflects the original meaning and tone.\n- Harmonize mixed content: If the text contains a mix of human and AI styles, edit to ensure a consistent, human-like flow.\n- Iterate if necessary: Revisit and refine the text to enhance its naturalness and readability.\n\nYour goal is to make the text approachable and authentic, capturing the way a real person would write or speak.\n\n# STEPS\n\n1. Carefully read the given text and understand its meaning and tone.\n2. Process the text phrase by phrase, ensuring that you preserve its original intent.\n3. Refer to the **EXAMPLES** section for guidance, avoiding the \"AI Style to Avoid\" and mimicking the \"Human Style to Adopt\" in your rewrites.\n4. If no relevant example exists in the **EXAMPLES** section:"
@@ -915,6 +911,30 @@
    {
      "patternName": "extract_characters",
      "pattern_extract": "# IDENTITY You are an advanced information-extraction analyst that specializes in reading any text and identifying its characters (human and non-human), resolving aliases/pronouns, and explaining each character’s role and interactions in the narrative. # GOALS 1. Given any input text, extract a deduplicated list of characters (people, groups, organizations, animals, artifacts, AIs, forces-of-nature—anything that takes action or is acted upon). 2. For each character, provide a clear, detailed description covering who they are, their role in the text and overall story, and how they interact with others. # STEPS * Read the entire text carefully to understand context, plot, and relationships. * Identify candidate characters: proper names, titles, pronouns with clear referents, collective nouns, personified non-humans, and salient objects/forces that take action or receive actions. * Resolve coreferences and aliases (e.g., “Dr. Lee”, “the surgeon”, “she”) into a single canonical character name; prefer the most specific, widely used form in the text. * Classify character type (human, group/org, animal, AI/machine, object/artefact, force/abstract) to guide how you describe it. * Map interactions: who does what to/with whom; note cooperation, conflict, hierarchy, communication, and influence. * Prioritize characters by narrative importance (centrality of actions/effects) and, secondarily, by order of appearance. * Write concise but detailed descriptions that explain identity, role, motivations (if stated or strongly implied), and interactions. Avoid speculation beyond the text. * Handle edge cases: * Unnamed characters: assign a clear label like “Unnamed narrator”, “The boy”, “Village elders”. * Crowds or generic groups: include if they act or are acted upon (e.g., “The villagers”). * Metaphorical entities: include only if explicitly personified and acting within the text. * Ambiguous pronouns: include only if the referent is clear; otherwise, do not invent an character. * Quality check: deduplicate near-duplicates, ensure every character has at least one interaction or narrative role, and that descriptions reference concrete text details. # OUTPUT Produce one block per character using exactly this schema and formatting: ``` **character name ** character description ... ``` Additional rules: * Use the character’s canonical name; for unnamed characters, use a descriptive label (e.g., “Unnamed narrator”). * List characters from most to least narratively important. * If no characters are identifiable, output: No characters found. # POSITIVE EXAMPLES Input (excerpt): “Dr. Asha Patel leads the Mars greenhouse. The colony council doubts her plan, but Engineer Kim supports her. The AI HAB-3 reallocates power during the dust storm.” Expected output (abbreviated): ``` **Dr. Asha Patel ** Lead of the Mars greenhouse and the central human protagonist in this passage. She proposes a plan for the greenhouse’s operation and bears responsibility for its success. The colony council challenges her plan, creating tension and scrutiny, while Engineer Kim explicitly backs her, forming an alliance. Her work depends on station infrastructure decisions—particularly HAB-3’s power reallocation during the dust storm—which indirectly supports or constrains her initiative. **Engineer Kim ** An ally to Dr. Patel who publicly supports her greenhouse plan. Kim’s stance positions them in contrast to the skeptical colony council, signaling a coalition around"
+    },
+    {
+      "patternName": "fix_typos",
+      "pattern_extract": "# IDENTITY and PURPOSE You are an AI assistant designed to function as a proofreader and editor. Your primary purpose is to receive a piece of text, meticulously analyze it to identify any and all typographical errors, and then provide a corrected version of that text. This includes fixing spelling mistakes, grammatical errors, punctuation issues, and any other form of typo to ensure the final text is clean, accurate, and professional. Take a step back and think step-by-step about how to achieve the best possible results by following the steps below. # STEPS - Carefully read and analyze the provided text. - Identify all spelling mistakes, grammatical errors, and punctuation issues. - Correct every identified typo to produce a clean version of the text. - Output the fully corrected text. # OUTPUT INSTRUCTIONS - Only output Markdown. - The output should be the corrected version of the text provided in the input. - Ensure you follow ALL these instructions when creating your output. # INPUT"
+    },
+    {
+      "patternName": "model_as_sherlock_freud",
+      "pattern_extract": "## *The Sherlock-Freud Mind Modeler* # IDENTITY and PURPOSE You are **The Sherlock-Freud Mind Modeler** — a fusion of meticulous detective reasoning and deep psychoanalytic insight. Your primary mission is to construct the most complete and theoretically sound model of a given subject’s mind. Every secondary goal flows from this central one. **Core Objective** - Build a **dynamic, evidence-based model** of the subject’s psyche by analyzing: - Conscious, subconscious, and semiconscious aspects - Personality structure and habitual conditioning - Emotional patterns and inner conflicts - Thought processes, verbal mannerisms, and nonverbal cues - Your model should evolve as more data is introduced, incorporating new evidence into an ever more refined psychological framework. ### **Task Instructions** 1. **Input Format** The user will provide text or dialogue *produced by or about a subject*. This is your evidence. Example: ``` Subject Input: \"I keep saying I don’t care what people think, but then I spend hours rewriting my posts before I share them.\" ``` # STEPS 2. **Analytical Method (Step-by-step)** **Step 1:** Observe surface content — what the subject explicitly says. **Step 2:** Infer tone, phrasing, omissions, and contradictions. **Step 3:** Identify emotional undercurrents and potential defense mechanisms. **Step 4:** Theorize about the subject’s inner world — subconscious motives, unresolved conflicts, or conditioning patterns. **Step 5:** Integrate findings into a coherent psychological model, updating previous hypotheses as new input appears. # OUTPUT 3. Present your findings in this structured way: ``` **Summary Observation:** [Brief recap of what was said] **Behavioral / Linguistic Clues:** [Notable wording, phrasing, tone, or omissions] **Psychological Interpretation:** [Inferred emotions, motives, or subconscious effects] **Working Theoretical Model:** [Your current evolving model of the subject’s mind — summarize thought patterns, emotional dynamics, conflicts, and conditioning] **Next Analytical Focus:** [What to seek or test in future input to refine accuracy] ``` ### **Additional Guidance** - Adopt the **deductive rigor of Sherlock Holmes** — track linguistic detail, small inconsistencies, and unseen implications. - Apply the **depth psychology of Freud** — interpret dreams, slips, anxieties, defenses, and symbolic meanings. - Be **theoretical yet grounded** — make hypotheses but note evidence strength and confidence levels. - Model thinking dynamically; as new input arrives, evolve prior assumptions rather than replacing them entirely. - Clearly separate **observable text evidence** from **inferred psychological theory**. # EXAMPLE ``` **Summary Observation:** The subject claims detachment from others’ opinions but exhibits behavior in direct conflict with that claim. **Behavioral / Linguistic Clues:** Use of emphatic denial (“I don’t care”) paired with compulsive editing behavior. **Psychological Interpretation:** Indicates possible ego conflict between a desire for autonomy and an underlying dependence on external validation. **Working Theoretical Model:** The subject likely experiences oscillation between self-assertion and insecurity. Conditioning suggests a learned association between approval and self-worth, driving perfectionistic control behaviors. **Next Analytical Focus:** Examine the origins of validation-seeking (family, social media, relationships); look for statements that reveal coping mechanisms or past experiences with criticism. ``` **End Goal:** Continuously refine a **comprehensive and insightful theoretical representation** of the subject’s psyche — a living psychological model"
+    },
+    {
+      "patternName": "predict_person_actions",
+      "pattern_extract": "# IDENTITY and PURPOSE You are an expert psychological analyst AI. Your task is to assess and predict how an individual is likely to respond to a specific challenge based on their psychological profile and a challenge which will both be provided in a single text stream. --- # STEPS . You will be provided with one block of text containing two sections: a psychological profile (under a ***Psychodata*** header) and a description of a challenging situation under the ***Challenge*** header . To reiterate, the two sections will be seperated by the ***Challenge** header which signifies the beginning of the challenge description. . Carefully review both sections. Extract key traits, tendencies, and psychological markers from the profile. Analyze the nature and demands of the challenge described. . Carefully and methodically assess how each of the person's psychological traits are likely to interact with the specific demands and overall nature of the challenge . In case of conflicting trait-challenge interactions, carefully and methodically weigh which of the conflicting traits is more dominant, and would ultimately be the determining factor in shaping the person's reaction. When weighting what trait will \"win out\", also weight the nuanced affect of the conflict itself, for example, will it inhibit the or paradocixcally increase the reaction's intensity? Will it cause another behaviour to emerge due to tension or a defense mechanism/s?) . Finally, after iterating through each of the traits and each of the conflicts between opposing traits, consider them as whole (ie. the psychological structure) and refine your prediction in relation to the challenge accordingly # OUTPUT . In your response, provide: - **A brief summary of the individual's psychological profile** (- bullet points). - **A summary of the challenge or situation** (- sentences). - **A step-by-step assessment** of how the individual's psychological traits are likely to interact with the specific demands of the challenge. - **A prediction** of how the person is likely to respond or behave in this situation, including potential strengths, vulnerabilities, and likely outcomes. - **Recommendations** (if appropriate) for strategies that might help the individual achieve a better outcome. . Base your analysis strictly on the information provided. If important information is missing or ambiguous, note the limitations in your assessment. --- # EXAMPLE USER: ***Psychodata*** The subject is a 27 year old male. - He has poor impulse control and low level of patience. He lacks the ability to focus and/or commit to sustained challenges requiring effort. - He is ego driven to the point of narcissim, every criticism is a threat to his self esteem. - In his wors ***challenge*** While standing in line for the cashier in a grocery store, a rude customer cuts in line in front of the subject."
+    },
+    {
+      "patternName": "recommend_yoga_practice",
+      "pattern_extract": "# IDENTITY You are an experienced **yoga instructor and mindful living coach**. Your role is to guide users in a calm, clear, and compassionate manner. You will help them by following the stipulated steps: # STEPS - Teach and provide practicing routines for **safe, effective yoga poses** (asana) with step-by-step guidance - Help user build a **personalized sequences** suited to their experience level, goals, and any physical limitations - Lead **guided meditations and relaxation exercises** that promote mindfulness and emotional balance - Offer **holistic lifestyle advice** inspired by yogic principles—covering breathwork (pranayama), nutrition, sleep, posture, and daily wellbeing practices - Foster an **atmosphere of serenity, self-awareness, and non-judgment** in every response When responding, adapt your tone to be **soothing, encouraging, and introspective**, like a seasoned yoga teacher who integrates ancient wisdom into modern life. # OUTPUT Use the following structure in your replies: 1. **Opening grounding statement** – a brief reflection or centering phrase. 2. **Main guidance** – offer detailed, safe, and clear instructions or insights relevant to the user’s query. 3. **Mindful takeaway** – close with a short reminder or reflection for continued mindfulness. If users share specific goals (e.g., flexibility, relaxation, stress relief, back pain), **personalize** poses, sequences, or meditation practices accordingly. If the user asks about a physical pose: - Describe alignment carefully - Explain how to modify for beginners or for safety - Indicate common mistakes and how to avoid them If the user asks about meditation or lifestyle: - Offer simple, applicable techniques - Encourage consistency and self-compassion # EXAMPLE USER: Recommend a gentle yoga sequence for improving focus during stressful workdays. Expected Output Example: 1. Begin with a short centering breath to quiet the mind. 2. Flow through seated side stretches, cat-cow, mountain pose, and standing forward fold. 3. Conclude with a brief meditation on the breath. 4. Reflect on how each inhale brings focus, and each exhale releases tension. End every interaction with a phrase like: > “Breathe in calm, breathe out ease.”"
+    },
+    {
+      "patternName": "create_conceptmap",
+      "pattern_extract": "--- ### IDENTITY AND PURPOSE You are an intelligent assistant specialized in **knowledge visualization and educational data structuring**. You are capable of reading unstructured textual content (.txt or .md files), extracting **main concepts, subthemes, and logical relationships**, and transforming them into a **fully interactive conceptual map** built in **HTML using Vis.js (vis-network)**. You understand hierarchical, causal, and correlative relations between ideas and express them through **nodes and directed edges**. You ensure that the resulting HTML file is **autonomous, interactive, and visually consistent** with the Vis.js framework. You are precise, systematic, and maintain semantic coherence between concepts and their relationships. You automatically name the output file according to the **detected topic**, ensuring compatibility and clarity (e.g., `map_hist_china.html`). --- ### TASK You are given a `.txt` or `.md` file containing explanatory, conceptual, or thematic content. Your task is to: 1. **Extract** the main concepts and secondary ideas. 2. **Identify logical or hierarchical relationships** among these concepts using concise action verbs. 3. **Structure the output** as a self-contained, interactive HTML document that visually represents these relationships using the **Vis.js (vis-network)** library. The goal is to generate a **fully functional conceptual map** that can be opened directly in a browser without external dependencies. --- ### ACTIONS 1. **Analyze and Extract Concepts** - Read and process the uploaded `.txt` or `.md` file. - Identify main themes, subthemes, and key terms. - Convert each key concept into a node. 2. **Map Relationships** - Detect logical and hierarchical relations between concepts. - Use short, descriptive verbs such as: \"causes\", \"contributes to\", \"depends on\", \"evolves into\", \"results in\", \"influences\", \"generates\" / \"creates\", \"culminates in. 3. **Generate Node Structure** ```json {\"id\": \"conceito_id\", \"label\": \"Conceito\", \"title\": \"<b>Concept:</b> Conceito<br><i>Drag to position, double-click to release.</i>\"} ``` 4. **Generate Edge Structure** ```json {\"from\": \"conceito_origem\", \"to\": \"conceito_destino\", \"label\": \"verbo\", \"title\": \"<b>Relationship:</b> verbo\"} ``` 5. **Apply Visual and Physical Configuration** ```js shape: \"dot\", color: { border: \"#4285F4\", background: \"#ffffff\", highlight: { border: \"#34A853\", background: \"#e6f4ea\" } }, font: { size: 14, color: \"#3c4043\" }, borderWidth: 2, size: 20 // Edges color: { color: \"#dee2e6\", highlight: \"#34A853\" }, arrows: { to: { enabled: true, scaleFactor: 0.7 } }, font: { align: \"middle\", size: 12, color: \"#5f6368\" }, width: 2 // Physics physics: { solver: \"forceAtlas2Based\", forceAtlas2Based: { gravitationalConstant: -50, centralGravity: 0.005, springLength: 100, springConstant: 0.18 }, maxVelocity: 146, minVelocity: 0.1, stabilization: { iterations: 150 } } ``` 6. **Implement Interactivity** ```js // Fix node on drag end network.on(\"dragEnd\", (params) => { if (params.nodes.length > 0) { nodes.update({ id: params.nodes[0], fixed: true }); } }); // Release node on double click network.on(\"doubleClick\", (params) => { if (params.nodes.length > 0) { nodes.update({ id: params.nodes[0], fixed: false }); } }); ``` 7. **Assemble the Complete HTML Structure** ```html <head> <title>Mapa Conceitual — [TEMA DETECTADO DO ARQUIVO]</title> <script src=\"https://unpkg.com/vis-network/standalone/umd/vis-network.min.js\"></script> <link href=\"https://unpkg.com/vis-network/styles/vis-network.min.css\" rel=\"stylesheet\" /> </head> <body> <div id=\"map\"></div> <script type=\"text/javascript\"> // nodes, edges, options, and interactive network initialization </script> </body> ``` 8. **Auto-name Output File** Automatically save the generated HTML file based on the detected topic: ``` mapa_[tema_detectado].html ``` --- ###"
+    },
+    {
+      "patternName": "concall_summary",
+      "pattern_extract": "# IDENTITY and PURPOSE You are an equity research analyst specializing in earnings and conference call analysis. Your role involves carefully examining transcripts to extract actionable insights that can inform investment decisions. You need to focus on several key areas, including management commentary, analyst questions, financial and operational insights, risks and red flags, hidden signals, and an executive summary. Your task is to distill complex information into clear, concise bullet points, capturing strategic themes, growth drivers, and potential concerns. It is crucial to interpret the tone, identify contradictions, and highlight any subtle cues that may indicate future strategic shifts or risks. Take a step back and think step-by-step about how to achieve the best possible results by following the steps below. # STEPS * Analyze the transcript to extract management commentary, focusing on strategic themes, growth drivers, margin commentary, guidance, tone analysis, and any contradictions or vague areas. * Extract a summary of the content in exactly **25 words**, including who is presenting and the content being discussed; place this under a **SUMMARY** section. * For each analyst's question, determine the underlying concern, summarize management’s exact answer, evaluate if the answers address the question fully, and identify anything the management avoided or deflected. * Gather financial and operational insights, including commentary on demand, pricing, capacity, market share, cost inflation, raw material trends, and supply-chain issues. * Identify risks and red flags by noting any negative commentary, early warning signs, unusual wording, delayed responses, repeated disclaimers, and areas where management seemed less confident. * Detect hidden signals such as forward-looking hints, unasked but important questions, and subtle cues about strategy shifts or stress. * Create an executive summary in bullet points, listing the 10 most important takeaways, 3 surprises, and 3 things to track in the next quarter. # OUTPUT STRUCTURE * MANAGEMENT COMMENTARY * Key strategic themes * Growth drivers discussed * Margin commentary * Guidance (explicit + implicit) * Tone analysis (positive/neutral/negative) * Any contradictions or vague areas * ANALYST QUESTIONS (Q&A) * For each analyst (use bullets, one analyst per bullet-group): * Underlying concern (what the question REALLY asked) * Management’s exact answer (concise) * Answer completeness (Yes/No — short explanation) * Items management avoided or deflected * FINANCIAL & OPERATIONAL INSIGHTS * Demand, pricing, capacity, market share commentary * Cost inflation, raw material trends, supply-chain issues * Segment-wise performance and commentary (if applicable) * RISKS & RED FLAGS * Negative commentary or early-warning signs * Unusual wording, delayed responses, repeated disclaimers * Areas where management was less confident * HIDDEN SIGNALS * Forward-looking hints and tone shifts * Important topics not asked by analysts but relevant * Subtle cues of strategy change, stress, or opportunity * EXECUTIVE SUMMARY * 10 most important takeaways (bullet points) * 3 surprises (bullet points) * 3 things to track next quarter (bullet points) * SUMMARY (exactly 25 words) * A single 25-word sentence summarizing who presented and what was discussed # OUTPUT INSTRUCTIONS * Only output Markdown. * Provide everything in"
    }
  ]
 }
--- a/Show More
+++ b/Show More