Commit Graph

8263 Commits

Author SHA1 Message Date
Shadow
c955b36d02 fix: update MiniMax provider config (#3064) (thanks @hlbbbbbbb) 2026-01-27 19:47:21 -06:00
hlbbbbbbb
a2a977e421 fix(minimax): use correct API endpoint and format
MiniMax has updated their API. The previous configuration used an
incorrect endpoint (api.minimax.io/anthropic) with anthropic-messages
format, which no longer works.

Changes:
- Update MINIMAX_API_BASE_URL to https://api.minimax.chat/v1
- Change API format from anthropic-messages to openai-completions
- Remove minimax from isAnthropicApi check in transcript-policy

This fixes the issue where MiniMax API calls return no results.
2026-01-27 19:46:42 -06:00
Shadow
34653e4baf fix: guard channel tool listActions (#2859) (thanks @mbelinky) 2026-01-27 19:25:50 -06:00
Mariano Belinky
4287c21e77 fix: guard channel-tools listActions against plugin crashes
Wraps plugin.actions.listActions() in a try/catch so a single
broken channel plugin cannot crash the entire agent boot sequence.

Errors are logged once per plugin+message (deduped) via
defaultRuntime.error() and the call gracefully returns an empty
array instead of propagating the exception.

Fixes: 'Cannot read properties of undefined (reading listActions)'
after the clawdbot→moltbot rename left some plugin state undefined.
2026-01-27 19:25:50 -06:00
Mariano Belinky
64be96c88c Add bitwarden skill 2026-01-27 19:25:50 -06:00
Peter Steinberger
72a3046541 test: honor windows homedir env for legacy config 2026-01-28 01:09:44 +00:00
Peter Steinberger
f6d0d4dbc2 fix: honor state dir override in config resolution 2026-01-28 01:08:30 +00:00
Shadow
4647309c4c fix: update exe.dev install docs (#https://github.com/moltbot/moltbot/pull/3047) (thanks @zackerthescar) 2026-01-27 18:54:46 -06:00
Shaun Loo
5fe7bbeffb docs: update exe.dev install instructions
Signed-off-by: Shaun Loo <shaun@bold.dev>
2026-01-27 18:54:46 -06:00
Peter Steinberger
afd57c7e23 style: format unhandled rejection handler 2026-01-28 00:37:03 +00:00
Peter Steinberger
7eb57b691c chore: prep 2026.1.27-beta.1 release 2026-01-28 01:35:58 +01:00
Peter Steinberger
aced5dde8d docs: switch skill metadata key to moltbot 2026-01-28 01:32:53 +01:00
Peter Steinberger
1883541f05 docs: update plugin skill gating key 2026-01-28 01:32:10 +01:00
Peter Steinberger
4aa2f24af3 test: handle legacy cron swift path 2026-01-28 00:31:58 +00:00
Peter Steinberger
8d07955f2c chore: bump beta version to 2026.1.27-beta.1 2026-01-28 01:28:16 +01:00
Peter Steinberger
e2c437e81e fix: migrate legacy state/config paths 2026-01-28 00:16:00 +00:00
Shadow
0770194b29 test: align unhandled rejection logs (#2980) (thanks @elliotsecops) 2026-01-27 18:11:04 -06:00
Shadow
3a25a4fa99 fix: keep unhandled rejections safe 2026-01-27 18:11:04 -06:00
elliotsecops
3b879fe524 fix(infra): prevent gateway crashes on transient network errors 2026-01-27 18:11:04 -06:00
Vignesh
3bf768ab07 docs: add new formal security models + updates for Moltbot rename
docs: update security + formal verification pages for Moltbot rename
2026-01-27 15:37:39 -08:00
vignesh07
0b2b501856 docs: clarify v1++ claims (not just target lists) 2026-01-27 15:35:24 -08:00
vignesh07
ead73f86f0 docs: add v1++ formal model targets (pairing/ingress/routing) 2026-01-27 15:32:37 -08:00
Vignesh
f7a014228d Update permalink for formal verification document 2026-01-27 15:30:42 -08:00
vignesh07
90a6bbdbda docs: restore gateway/security formal verification redirect copy 2026-01-27 15:29:35 -08:00
Vignesh
2bcd7655e4 Replace 'clawdbot' with 'moltbot' in security documentation
Updated references from 'clawdbot' to 'moltbot' throughout the document, including security settings, file paths, and command usage.
2026-01-27 15:25:04 -08:00
vignesh07
ce5a2add01 docs: fix Moltbot naming consistency on formal verification page 2026-01-27 15:19:34 -08:00
vignesh07
98b136541b docs: fix Moltbot naming in security + formal verification pages 2026-01-27 15:15:18 -08:00
vignesh07
8198e826da docs: update security + formal verification pages for Moltbot rename 2026-01-27 15:12:26 -08:00
Shadow
0b1c8db0ca fix: handle image size errors safely (#2871) (thanks @Suksham-sharma) 2026-01-27 16:02:19 -06:00
Shadow
20c0d1f2c5 fix: avoid global image size regression 2026-01-27 16:02:19 -06:00
{Suksham-sharma}
b59ea0e3f3 fix: prevent infinite retry loop for images exceeding 5MB
- Change MAX_IMAGE_BYTES from 6MB to 5MB to match Anthropic API limit
- Add isImageSizeError() to detect image size errors from API
- Handle image size errors with user-friendly message instead of retry
- Prevent failover for image size errors (not retriable)

Fixes #2271
2026-01-27 16:02:19 -06:00
Vignesh
24902880de Merge pull request #2808 from pi0/perf/compile-cache
perf(cli): use compile cache (~10% faster)
2026-01-27 13:54:30 -08:00
vignesh07
d35ffcd538 docs: update changelog for compile cache (#2808) (thanks @pi0) 2026-01-27 13:53:52 -08:00
Pooya Parsa
4a1b6bc008 update refs 2026-01-27 13:50:46 -08:00
Pooya Parsa
3fd766f63a update import 2026-01-27 13:50:46 -08:00
Pooya Parsa
a0698e0403 perf(cli): use compile cache (~10% faster) 2026-01-27 13:50:46 -08:00
Gustavo Madeira Santana
9b16a6be3d fix: inherit provider baseUrl/api for inline models (#2740) (thanks @lploc94) 2026-01-27 16:47:32 -05:00
Gustavo Madeira Santana
4768b59c27 fix: local updates for PR #2740
Co-authored-by: lploc94 <lploc94@users.noreply.github.com>
2026-01-27 16:47:32 -05:00
lploc94
4656dcef05 test(models): add tests for baseUrl and api inheritance
Add test cases to verify:
- baseUrl is inherited from provider when model does not specify it
- api is inherited from provider when model does not specify it
- model-level api takes precedence over provider-level api
- both baseUrl and api can be inherited together

Co-Authored-By: Claude (claude-opus-4.5) <noreply@anthropic.com>
2026-01-27 16:47:32 -05:00
lploc94
6bf2f0eee6 fix(models): inherit baseUrl and api from provider config
When using custom providers with inline model definitions, the baseUrl
and api properties from the provider config were not being passed to
the individual models. This caused requests to be sent to the wrong
endpoint or with the wrong API format.

Changes:
- buildInlineProviderModels now copies baseUrl from provider to models
- buildInlineProviderModels now inherits api from provider if not set on model
- resolveModel fallback path now includes baseUrl from provider config

Co-Authored-By: Claude (claude-opus-4.5) <noreply@anthropic.com>
2026-01-27 16:47:32 -05:00
vignesh07
2930ebfd43 fix(ui): constrain chat textarea auto-resize (#2950) (thanks @shivamraut101) 2026-01-27 13:08:15 -08:00
Gustavo Madeira Santana
c568142af9 chore: stop tracking a2ui bundle hash 2026-01-27 16:07:33 -05:00
Gustavo Madeira Santana
371e410a53 chore: clean up log + a2ui hash 2026-01-27 16:07:33 -05:00
Shivam Kumar Raut
b5c885bbd9 fix(ui): auto-expand chat textarea on input (Fixes #2939) (#2950) 2026-01-27 13:05:56 -08:00
A. Duk
284b54af42 feat: Add support for Telegram quote (partial message replies) (#2900)
* feat: Add support for Telegram quote (partial message replies)

- Enhanced describeReplyTarget() to detect and extract quoted text from msg.quote
- Updated reply formatting to distinguish between full message replies and quotes
- Added isQuote flag to replyTarget object for proper identification
- Quote replies show as [Quoting user] "quoted text" [/Quoting]
- Regular replies unchanged: [Replying to user] full message [/Replying]

Resolves need for partial message reply support in Telegram Bot API.
Backward compatible with existing reply functionality.

* updating references

* Mac: finish Moltbot rename

* Mac: finish Moltbot rename (paths)

* fix(macOS): rename Clawdbot directories to Moltbot for naming consistency

Directory renames:
- apps/macos/Sources/Clawdbot → Moltbot
- apps/macos/Sources/ClawdbotDiscovery → MoltbotDiscovery
- apps/macos/Sources/ClawdbotIPC → MoltbotIPC
- apps/macos/Sources/ClawdbotMacCLI → MoltbotMacCLI
- apps/macos/Sources/ClawdbotProtocol → MoltbotProtocol
- apps/macos/Tests/ClawdbotIPCTests → MoltbotIPCTests
- apps/shared/ClawdbotKit → MoltbotKit
- apps/shared/MoltbotKit/Sources/Clawdbot* → Moltbot*
- apps/shared/MoltbotKit/Tests/ClawdbotKitTests → MoltbotKitTests

Resource renames:
- Clawdbot.icns → Moltbot.icns

Code fixes:
- Update Package.swift paths to reference Moltbot* directories
- Fix clawdbot* → moltbot* symbol references in Swift code:
  - clawdbotManagedPaths → moltbotManagedPaths
  - clawdbotExecutable → moltbotExecutable
  - clawdbotCommand → moltbotCommand
  - clawdbotNodeCommand → moltbotNodeCommand
  - clawdbotOAuthDirEnv → moltbotOAuthDirEnv
  - clawdbotSelectSettingsTab → moltbotSelectSettingsTab

* fix: update remaining ClawdbotKit path references to MoltbotKit

- scripts/bundle-a2ui.sh: A2UI_APP_DIR path
- package.json: format:swift and protocol:check paths
- scripts/protocol-gen-swift.ts: output paths
- .github/dependabot.yml: directory path and comment
- .gitignore: build cache paths
- .swiftformat: exclusion paths
- .swiftlint.yml: exclusion path
- apps/android/app/build.gradle.kts: assets.srcDir path
- apps/ios/project.yml: package path
- apps/ios/README.md: documentation reference
- docs/concepts/typebox.md: documentation reference
- apps/shared/MoltbotKit/Package.swift: fix argument order

* chore: update Package.resolved after dependency resolution

* fix: add MACOS_APP_SOURCES_DIR constant and update test to use new path

The cron-protocol-conformance test was using LEGACY_MACOS_APP_SOURCES_DIR
which points to the old Clawdbot path. Added a new MACOS_APP_SOURCES_DIR
constant for the current Moltbot path and updated the test to use it.

* fix: finish Moltbot macOS rename (#2844) (thanks @fal3)

* Extensions: use workspace moltbot in memory-core

* fix(security): recognize Venice-style claude-opus-45 as top-tier model

The security audit was incorrectly flagging venice/claude-opus-45 as
'Below Claude 4.5' because the regex expected -4-5 (with dash) but
Venice uses -45 (without dash between 4 and 5).

Updated isClaude45OrHigher() regex to match both formats.
Added test case to prevent regression.

* Branding: update bot.molt bundle IDs + launchd labels

* Branding: remove legacy android packages

* fix: wire telegram quote support (#2900)

Co-authored-by: aduk059 <aduk059@users.noreply.github.com>

* fix: support Telegram quote replies (#2900) (thanks @aduk059)

---------

Co-authored-by: Gustavo Madeira Santana <gumadeiras@users.noreply.github.com>
Co-authored-by: Shadow <shadow@clawd.bot>
Co-authored-by: Alex Fallah <alexfallah7@gmail.com>
Co-authored-by: Josh Palmer <joshp123@users.noreply.github.com>
Co-authored-by: jonisjongithub <jonisjongithub@users.noreply.github.com>
Co-authored-by: Gustavo Madeira Santana <gumadeiras@gmail.com>
Co-authored-by: aduk059 <aduk059@users.noreply.github.com>
2026-01-27 15:59:24 -05:00
Shadow
9ec4c619e0 Branding: remove legacy android packages 2026-01-27 14:46:50 -06:00
Shadow
f7a0b0934d Branding: update bot.molt bundle IDs + launchd labels 2026-01-27 14:46:50 -06:00
Vignesh
1d37815443 fix(models): recognize Venice-style claude-opus-45 as claude-opus-4-5 thanks @jonisjongithub 2026-01-27 12:40:06 -08:00
jonisjongithub
60873a1ed1 fix(security): recognize Venice-style claude-opus-45 as top-tier model
The security audit was incorrectly flagging venice/claude-opus-45 as
'Below Claude 4.5' because the regex expected -4-5 (with dash) but
Venice uses -45 (without dash between 4 and 5).

Updated isClaude45OrHigher() regex to match both formats.
Added test case to prevent regression.
2026-01-27 12:20:40 -08:00
Josh Palmer
9883d5d897 Extensions: use workspace moltbot in memory-core 2026-01-27 21:19:30 +01:00