Add new AutoGPT Platform Block that uses google/gemini-2.5-flash-image
model via Replicate API.
Features:
- Text prompt input for image generation
- Optional list of image URLs as input
- Configurable output format (jpg/png, defaults to png)
- Single model option: google/gemini-2.5-flash-image
- Returns image_url output for generated images
Fixes#10815🤖 Generated with [Claude Code](https://claude.ai/code)
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] use the AI image customizer block and upload 2 images to see if it
uses them in the image generation/edits
<img width="1536" height="672" alt="tmprhzqasxz"
src="https://github.com/user-attachments/assets/39d7adbd-2847-4988-aeab-1c5453290174"
/>
---------
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Swifty <craigswift13@gmail.com>
Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com>
Co-authored-by: Bently <Bentlybro@users.noreply.github.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Adds support for Ideogram V3 model while maintaining backward
compatibility with existing models (V1,
V1_TURBO, V2, V2_TURBO). Updates default model to V3 and implements
smart API routing to handle
Ideogram's new V3 endpoint requirements.
Changes Made
- Added V3 model support: Added V_3 to IdeogramModelName enum and set as
default
- Dual API endpoint handling:
- V3 models route to new /v1/ideogram-v3/generate endpoint with updated
payload format
- Legacy models (V1, V2, Turbo variants) continue using /generate
endpoint
- Model-specific feature filtering:
- V1 models: Basic parameters only (no style_type or color_palette
support)
- V2/V2_TURBO: Full legacy feature support including style_type and
color_palette
- V3: New endpoint with aspect ratio mapping and updated parameter
structure
- Aspect ratio compatibility: Added mapping between internal enum values
and V3's expected format
(ASPECT_1_1 → 1x1)
- Updated pricing: V3 model costs 18 credits (vs 16 for other models)
- Updated default usage: Store image generation now uses V3 by default
Technical Details
Ideogram updated their API with a separate V3 endpoint that has
different requirements:
- Different URL path (/v1/ideogram-v3/generate)
- Different aspect ratio format (e.g., 1x1 instead of ASPECT_1_1)
- Model-specific feature support (V1 models don't support style_type,
etc.)
The implementation intelligently routes requests to the appropriate
endpoint based on the selected model
while maintaining a single unified interface.
I tested all the models and they are working here
<img width="1804" height="887" alt="image"
src="https://github.com/user-attachments/assets/9f2e44ca-50a4-487f-987c-3230dd72fb5e"
/>
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Test the Ideogram model block and watch as they all work!
Added basic stagehand integration:
<img width="667" height="609" alt="Screenshot 2025-08-27 at 09 20 18"
src="https://github.com/user-attachments/assets/11ab2941-0913-4346-a1d4-45980711e0f9"
/>
[stagehand_v35.json](https://github.com/user-attachments/files/22002924/stagehand_v35.json)
### Changes 🏗️
- Act Block
- Extract Block
- Observe Block
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] I have added a sample agent
- [x] I have created an agent that uses these blocks and ensured it runs
### Changes 🏗️
- Updated the creator page to utilize React Query for data fetching,
improving performance and reliability.
- Removed legacy API calls and integrated prefetching for creator
details and agents.
- Introduced a new MainCreatorPage component for better separation of
concerns.
- Added a hydration boundary for managing server state.
### Checklist 📋
### Checklist 📋
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] All marketplace E2E tests are working.
- [x] I’ve tested all the links and checked if everything renders
perfectly on the marketplace page.
- resolves -
https://github.com/Significant-Gravitas/AutoGPT/issues/10618
When we have a dropdown with a large description, the actions button is
moved out of the dialog box. To fix this, I’ve added a temporary
solution, but in the future, we need to change the entire layout.
### Checklist 📋
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Everything works perfectly locally.
- Fixes#10749
### Changes 🏗️
- Fix implementation of `useAgentRunsInfinite.upsertAgentRun(..)`
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] New runs appear in runs list
### Changes 🏗️
This PR fixes an infinite loop issue in the execution manager where
malformed or unparseable messages would be continuously requeued,
causing high CPU usage and preventing the system from processing
legitimate messages.
**Key changes:**
- Modified `_ack_message()` function to accept explicit `requeue`
parameter
- Set `requeue=False` for malformed/unparseable messages that cannot be
fixed by retrying
- Set `requeue=False` for duplicate execution attempts (graph already
running)
- Kept `requeue=True` for legitimate failures that may succeed on retry
(e.g., temporary resource constraints, network issues)
**Technical details:**
The previous implementation always set `requeue=True` when rejecting
messages with `basic_nack()`. This caused problematic messages to be
immediately re-delivered to the consumer, creating an infinite loop for:
1. Messages with invalid JSON that cannot be parsed
2. Messages for executions that are already running (duplicates)
These scenarios will never succeed regardless of how many times they're
retried, so they should be rejected without requeueing to prevent
resource exhaustion.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Verified malformed messages are rejected without requeue
- [x] Confirmed duplicate execution messages are rejected without
requeue
- [x] Ensured legitimate failures (shutdown, pool full) still requeue
properly
- [x] Tested that normal message processing continues to work correctly
Backend for the Blocks Menu Redesign.
### Changes 🏗️
- Add optional `agent_name` to the `AgentExecutorBlock` - displayed as
the block name in the Builder
- Include `output_schema` in the `LibraryAgent` model
- Make `v2.store.db.py:get_store_agents` accept multiple creators filter
- Add `api/builder` router with endpoints (and accompanying logic in
`v2/builder/db` and models in `v2/builder/models`)
- `/suggestions`: elements for the suggestions tab
- `/categories`: categories with a number of blocks per each
- `/blocks`: blocks based on category, type or provider
- `/providers`: integration providers with their block counts
- `/serach`: search blocks (including integrations), marketplace agents
and user library agents
- `/counts`: element counts for each category in the Blocks Menu.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Modified function `get_store_agents` works in existing code paths
- [x] Agent executor block works
- [x] New endpoints work
- [x] Existing Builder menu is unaffected
---------
Co-authored-by: Abhimanyu Yadav <abhimanyu1992002@gmail.com>
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
- Resolves#10713
### Changes 🏗️
- Remove early exit in API proxy that suppresses auth errors
- Remove unused `proxy-action.ts`
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Publish Agent dialog works when logged out
- [x] Publish Agent dialog works when logged in
---------
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
The database manager had both sync and async clients that contained
overlapping methods, including some that weren't actually used in their
respective contexts. This violated the principle that each client should
only expose the methods it needs.
## Problem
The `DatabaseManagerClient` (sync) included
`get_user_execution_summary_data`, but this method was only ever used in
async contexts like the notifications system. This created unnecessary
coupling and violated the design goal of having focused,
context-specific clients.
## Solution
After comprehensive analysis of actual method usage across the codebase:
- **Removed** `get_user_execution_summary_data` from
`DatabaseManagerClient` since it's only used in async contexts
(notifications)
- **Verified** all remaining methods on both clients are actively used
in their respective contexts:
- Sync client (11 methods): Used in monitoring and main execution thread
- Async client (26 methods): Used in node execution, blocks, and
notifications
- **Maintained** the base `DatabaseManager` class with the union of all
methods needed by either client
## Impact
Each client now contains exactly the methods it needs for its specific
usage patterns:
- `DatabaseManagerClient` handles synchronous operations like monitoring
and credit management
- `DatabaseManagerAsyncClient` handles asynchronous operations like node
execution, persistence, and notifications
The change is minimal and surgical - only removing one unused method
while preserving all actually-used functionality.
Fixes#10658.
<!-- START COPILOT CODING AGENT TIPS -->
---
✨ Let Copilot coding agent [set things up for
you](https://github.com/Significant-Gravitas/AutoGPT/issues/new?title=✨+Set+up+Copilot+instructions&body=Configure%20instructions%20for%20this%20repository%20as%20documented%20in%20%5BBest%20practices%20for%20Copilot%20coding%20agent%20in%20your%20repository%5D%28https://gh.io/copilot-coding-agent-tips%29%2E%0A%0A%3COnboard%20this%20repo%3E&assignees=copilot)
— coding agent works faster and does higher quality work when set up for
your repo.
---------
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Swifty <craigswift13@gmail.com>
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: ntindle <8845353+ntindle@users.noreply.github.com>
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Nicholas Tindle <nicktindle@outlook.com>
Co-authored-by: Bently <Github@bentlybro.com>
Handle invalid or empty response from MusicGen model
Fixes: #9145
> ⚠️ Note: This PR does not directly fix issue #9145 (failed run marked
as success), but improves the validation of the URL to reduce the
chances of invalid states entering the system. This is a related
improvement, but not the root cause fix.
### Description
During execution of the meta/musicgen model via Replicate API, the
application failed
with an error indicating the model returned an empty or invalid
response.
Although some API calls succeeded, this error showed the logic was not
checking the
structure and content of the result properly before processing it.
PROBLEM:
CONTEXT:
API: Replicate
MODEL: meta/musicgen:671ac645
STATUS: Failed after 3 attempts
ERROR_MESSAGE: "Unexpected error: Model returned empty or invalid
response"
CAUSE:
- The original logic did not validate result structure.
- It assumed any non-null output was valid, including strings like "No
output received".
- This led to invalid/malformed results being passed to the frontend.
### Changes 🏗️
- Added `AIMusicGeneratorBlock` to support music generation using Meta’s
MusicGen models via Replicate API.
- Supports configurable inputs like prompt, model version, duration,
temperature, top_k/p, and normalization.
- Uses robust retry logic for reliability.
- Output returns audio URL; errors return user-friendly message.
BEFORE_CODE: |
```
if result and result != "No output received":
yield "result", result
return
```
AFTER_CODE: |
```
if result and isinstance(result, str) and result.startswith("http"):
yield "result", result
return
```
### Checklist 📋
#### For code changes:
- [x] Clearly listed changes in the PR description
- [x] Added test plan and mock outputs
- [x] Tested with various prompts and confirmed working output
### Test Plan
- [x] Ran locally with valid Replicate API key
- [x] Generated audio with different prompts
- [x] Simulated failure to verify retry and error message
---------
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
this fixes and makes the moderation message properly show the moderation
ID
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] try trigger moderation and have it shows the moderation id in the
error message
### Changes 🏗️
This PR implements email notifications for agent creators when their
agent submissions are approved or rejected by an admin in the
marketplace.
Specifically, the changes include:
- Added `AGENT_APPROVED` and `AGENT_REJECTED` notification types to
`schema.prisma`.
- Created `AgentApprovalData` and `AgentRejectionData` Pydantic models
for notification data.
- Configured the notification system to use immediate queues and new
Jinja2 templates for these types.
- Designed two new email templates: `agent_approved.html.jinja2` and
`agent_rejected.html.jinja2`, with dynamic content for agent details,
reviewer feedback, and relevant action links.
- Modified the `review_store_submission` function to:
- Include `User` and `Reviewer` data in the database query.
- Construct and queue the appropriate email notification based on the
approval/rejection status.
- Ensure email sending failures do not block the agent review process.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Approve an agent via the admin dashboard.
- [x] Verify the agent creator receives an "Agent Approved" email with
correct details and a link to the store.
- [x] Reject an agent via the admin dashboard (providing a reason).
- [x] Verify the agent creator receives an "Agent Rejected" email with
correct details, the rejection reason, and a link to resubmit.
- [x] Verify that if email sending fails (e.g., misconfigured SMTP), the
agent approval/rejection process still completes successfully without
error.
<img width="664" height="975" alt="image"
src="https://github.com/user-attachments/assets/d397f2dc-56eb-45ab-877e-b17f1fc234d1"
/>
<img width="664" height="975" alt="image"
src="https://github.com/user-attachments/assets/25597752-f68c-46fe-8888-6c32f5dada01"
/>
---
Linear Issue: [SECRT-1168](https://linear.app/autogpt/issue/SECRT-1168)
<a
href="https://cursor.com/background-agent?bcId=bc-7394906c-0341-4bd0-8842-6d9d6f83c56c">
<picture>
<source media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/open-in-cursor-dark.svg">
<source media="(prefers-color-scheme: light)"
srcset="https://cursor.com/open-in-cursor-light.svg">
<img alt="Open in Cursor" src="https://cursor.com/open-in-cursor.svg">
</picture>
</a>
<a
href="https://cursor.com/agents?id=bc-7394906c-0341-4bd0-8842-6d9d6f83c56c">
<picture>
<source media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/open-in-web-dark.svg">
<source media="(prefers-color-scheme: light)"
srcset="https://cursor.com/open-in-web-light.svg">
<img alt="Open in Web" src="https://cursor.com/open-in-web.svg">
</picture>
</a>
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
AutoModManager now captures and propagates the content_id from the
moderation API for both input and output moderation. AutoModResponse and
ModerationError are updated to include content_id, allowing better
traceability of moderation actions and error reporting, with this the
error message will now show ``Failed due to content moderation
(Moderation ID: uuid-here)``
This is good for if a user is having a issue with automod and its
falsely flagging there runs we can use the moderation ID to look at
automod to see whats going on
<!-- Clearly explain the need for these changes: -->
### Changes 🏗️
Just some updates to receive the content_id from AutoMod and then show
it in the "Failed due to content moderation" message
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Run autogpt with AM and trigger it and it will show the content_id
in the error message
<!-- Clearly explain the need for these changes: -->
### Need for these changes 💥
This PR resolves Linear issue `SECRT-1290`, addressing a critical bug
where the scheduler API fails with a "Wrong number of fields" error when
empty or invalid cron expressions are submitted from the frontend. This
was causing production errors and a poor user experience. It was an off
by one error
### Changes 🏗️
Fix off by one error + add additional logging / error messaging when
someone makes an invalid cron
https://github.com/user-attachments/assets/775881a9-707b-4c4f-b23a-bd7118a358ee
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Attempt to schedule an agent with an empty cron expression from
the UI and confirm a frontend toast error.
- [x] Attempt to schedule an agent with an incomplete yearly cron (no
months selected) from the UI and confirm a frontend toast error and UI
warning.
- [x] Attempt to schedule an agent with an incomplete monthly cron (no
days selected) from the UI and confirm a frontend toast error and UI
warning.
- [x] Attempt to schedule an agent with an incomplete weekly cron (no
days selected) from the UI and confirm a frontend toast error and UI
warning.
- [x] Verify that valid cron expressions can still be scheduled
successfully.
- [x] Run backend unit tests for scheduler cron validation.
- [x] Run frontend unit tests for cron expression utility.
---
Linear Issue: [SECRT-1290](https://linear.app/autogpt/issue/SECRT-1290)
<a
href="https://cursor.com/background-agent?bcId=bc-8bc10502-9498-4dbd-afa2-93e15990fa8c">
<picture>
<source media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/open-in-cursor-dark.svg">
<source media="(prefers-color-scheme: light)"
srcset="https://cursor.com/open-in-cursor-light.svg">
<img alt="Open in Cursor" src="https://cursor.com/open-in-cursor.svg">
</picture>
</a>
<a
href="https://cursor.com/agents?id=bc-8bc10502-9498-4dbd-afa2-93e15990fa8c">
<picture>
<source media="(prefers-color-scheme: dark)"
srcset="https://cursor.com/open-in-web-dark.svg">
<source media="(prefers-color-scheme: light)"
srcset="https://cursor.com/open-in-web-light.svg">
<img alt="Open in Web" src="https://cursor.com/open-in-web.svg">
</picture>
</a>
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Swifty <craigswift13@gmail.com>
At the bottom of the library page is an alert directing people to the
old monitoring page. This PR removes this link, as the old monitoring
page no longer works.
### Changes 🏗️
Remove Alert directing users to the old monitoring page.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Check alert no longer shows on the library page
<!-- Clearly explain the need for these changes: -->
This PR implements blocks that enable users to interact with the AutoGPT
store and library programmatically. This addresses the need for agents
to be able to add other agents from the store to their library and
manage agent collections automatically, as requested in Linear issue
OPEN-2602. These are locked behind LaunchDarkly for now.
https://github.com/user-attachments/assets/b8518961-abbf-4e9d-a31e-2f3d13fa6b0d
### Changes 🏗️
<!-- Concisely describe all of the changes made in this pull request:
-->
- **Added new store operations blocks**
(`backend/blocks/system/store_operations.py`):
- `GetStoreAgentDetailsBlock`: Retrieves detailed information about an
agent from the store
- `SearchStoreAgentsBlock`: Searches for agents in the store with
various filters
- **Added new library operations blocks**
(`backend/blocks/system/library_operations.py`):
- `ListLibraryAgentsBlock`: Lists all agents in the user's library
- `AddToLibraryFromStoreBlock`: Adds an agent from the store to user's
library
- **Updated block exports** in `backend/blocks/system/__init__.py` to
include new blocks
- **Added comprehensive tests** for store operations in
`backend/blocks/test/test_store_operations.py`
- **Enhanced executor database utilities** in
`backend/executor/database.py` with new helper methods for agent
management
- **Updated frontend marketplace page** to properly handle the new store
operations
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Created unit tests for all new store operation blocks
- [x] Tested GetStoreAgentDetailsBlock retrieves correct agent
information
- [x] Tested SearchStoreAgentsBlock filters and returns agents correctly
- [x] Tested AddToLibraryFromStoreBlock successfully adds agents to
library
- [x] Tested error handling for non-existent agents and invalid inputs
- [x] Verified all blocks integrate properly with the database manager
- [x] Confirmed blocks appear in the block registry and are accessible
---------
Co-authored-by: Lluis Agusti <hi@llu.lu>
Co-authored-by: Swifty <craigswift13@gmail.com>
This simply removes the schedule "every minute" option from the schedule
tasks UI, Its still possible to set a every minute schedule via the
custom option
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Check the Schedule Tasks UI to see there is no more "Every Minute"
option
- [x] Check you can still set a schedule a agent to run every minute
using the custom option
<!-- Clearly explain the need for these changes: -->
We're showing invalid credential types on the selections across the app
### Changes 🏗️
We go from (incorrect)
<img width="2551" height="1202" alt="image"
src="https://github.com/user-attachments/assets/e566ed6c-b6c9-4047-80fd-0f2c8cef0bf9"
/>
to this with the fix
<img width="2551" height="1202" alt="image"
src="https://github.com/user-attachments/assets/c720a3d4-9c03-48c5-82a3-d30752bce13c"
/>
<!-- Concisely describe all of the changes made in this pull request:
-->
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] test the broken agent and upload images proving its no longer
broken
<!-- Clearly explain the need for these changes: -->
We want a way to get the user's id from discord without them having to
enable dev mode so this is a way -- oauth login
<img width="2551" height="1202" alt="image"
src="https://github.com/user-attachments/assets/71be07a9-fd37-4ea7-91a1-ced8972fda29"
/>
### Changes 🏗️
- Created DiscordOAuthHandler for managing OAuth2 flow, including login
URL generation, token exchange, and revocation.
- Implemented support for PKCE in the OAuth2 flow.
- Enhanced error handling for user info retrieval and token management.
- Add discord block for getting the logged in user
- Add new client secret field to .env.default
<!-- Concisely describe all of the changes made in this pull request:
-->
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] add the blocks and test they all work
#### For configuration changes:
- [x] `.env.default` is updated or already compatible with my changes
- [x] `docker-compose.yml` is updated or already compatible with my
changes
- [x] I have included a list of my configuration changes in the PR
description (under **Changes**)
<!-- Clearly explain the need for these changes: -->
We need the ability to reject or remove already approved agents from the
marketplace via the Admin Dashboard. Previously, once an agent was
approved, there was no easy way to remove it from the marketplace
without direct database intervention.
This addresses several use cases:
- Removing agents that require credentials (short-term solution
discussed with Reinier)
- Handling broken agents mistakenly approved
- Managing outdated or problematic agents
- Quick response to issues without engineering support
### Changes 🏗️
<!-- Concisely describe all of the changes made in this pull request:
-->
- **Backend**: Modified `review_store_submission` function in
`/backend/server/v2/store/db.py` to handle rejecting already approved
agents
- Added logic to detect when rejecting an approved agent
- Updates StoreListing to remove agent from marketplace when rejected
- Handles multiple approved versions correctly
- **Frontend**: Updated admin marketplace UI components
- `expandable-row.tsx`: Show action buttons for both PENDING and
APPROVED agents
- `approve-reject-buttons.tsx`:
- Show only "Revoke" button for approved agents (hide Approve button)
- Update button text from "Reject" to "Revoke" for approved agents
- Update dialog titles and descriptions appropriately
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Navigate to Admin Dashboard > Marketplace Management
- [x] Find a PENDING agent and verify both Approve and Reject buttons
appear
- [x] Find an APPROVED agent and verify only "Revoke" button appears
- [x] Click Revoke on an approved agent and verify dialog shows "Revoke
Approved Agent" title
- [x] Submit revocation with comments and verify agent status changes to
REJECTED
- [x] Verify the agent is removed from the public marketplace
- [x] Test with an agent that has multiple approved versions - verify it
switches to another approved version
- [x] Test with an agent that has only one approved version - verify
hasApprovedVersion is set to false
#### For configuration changes:
- [x] `.env.example` is updated or already compatible with my changes
- [x] `docker-compose.yml` is updated or already compatible with my
changes
- [x] I have included a list of my configuration changes in the PR
description (under **Changes**)
No configuration changes required - this uses existing admin
authentication and database schema.
Fixes SECRT-1218
🤖 Generated with [Claude Code](https://claude.ai/code)
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
- Resolves#10645
### Changes 🏗️
- Implement infinite scroll in the Agent Runs list (on
`/library/agents/[id]`)
- Add horizontal scroll support to `ScrollArea` and `InfiniteScroll`
components
- Fix `InfiniteScroll` triggering twice
- Fix date handling by React Queries
- Add response mutator to parse dates coming out of API
- Make legacy `GraphExecutionMeta` compatible with generated type
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- Open `/library/agents/[id]`
- [x] Agent runs list loads
- Scroll agent runs list to the end
- [x] More runs are loaded and appear in the list
### Changes 🏗️
This PR adds Meeting BaaS (Bot-as-a-Service) integration to the AutoGPT
platform, enabling automated meeting recording and transcription
capabilities.
<img width="1157" height="633" alt="Screenshot 2025-08-22 at 15 06 15"
src="https://github.com/user-attachments/assets/53b88bc8-5580-4287-b6ed-3ae249aed69f"
/>
[BAAS
Test_v12.json](https://github.com/user-attachments/files/21938290/BAAS.Test_v12.json)
**New Features:**
- **Meeting Recording Bot Management:**
- Deploy bots to join and record meetings automatically
- Support for multiple meeting platforms via meeting URL
- Scheduled bot deployment with Unix timestamp support
- Custom bot avatars and entry messages
- Webhook support for real-time event notifications
- **Meeting Data Operations:**
- Retrieve MP4 recordings and transcripts from completed meetings
- Delete recording data for privacy/storage management
- Force bots to leave ongoing meetings
**Technical Implementation:**
- Added 4 new files under `backend/blocks/baas/`:
- `__init__.py`: Package initialization
- `_api.py`: Meeting BaaS API client with comprehensive endpoints
- `_config.py`: Provider configuration using SDK pattern
- `bots.py`: 4 bot management blocks (Join, Leave, Fetch Data, Delete
Recording)
**Key Capabilities:**
- Join meetings with customizable bot names and avatars
- Automatic transcription with configurable speech-to-text providers
- Time-limited MP4 download URLs for recordings
- Reserved bot slots for joining 4 minutes before meetings
- Automatic leave timeouts configuration
- Custom metadata support for tracking
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Created and executed an agent with Meeting BaaS bot deployment
blocks
- [x] Tested bot joining meetings with various configurations
- [x] Verified recording retrieval and transcript functionality
- [x] Tested bot removal from ongoing meetings
- [x] Confirmed data deletion operations work correctly
- [x] Verified error handling for invalid API keys and bot IDs
- [x] Tested webhook URL configuration
---------
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
### Changes 🏗️
Expanded the Ollama setup instructions to include desktop app, Docker,
and legacy CLI methods. Added new screenshots for network exposure and
model selection. Clarified steps for starting the AutoGPT platform,
configuring models, and troubleshooting. Included instructions for
adding custom models and improved overall documentation structure.
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Look at the new ollama doc to make sure it makes sense and is easy
to follow
<!-- Clearly explain the need for these changes: -->
This PR adds two new DataForSEO blocks for keyword research
functionality, enabling users to get keyword suggestions and related
keywords using the DataForSEO Labs API.
https://github.com/user-attachments/assets/55b3f64b-20b2-4c6d-b307-01327d476fe2
[DataForSeo
Poc_v3.json](https://github.com/user-attachments/files/21916605/DataForSeo.Poc_v3.json)
### Changes 🏗️
<!-- Concisely describe all of the changes made in this pull request:
-->
- Added `DataForSeoKeywordSuggestionsBlock` for getting keyword
suggestions from DataForSEO Labs
- Added `DataForSeoRelatedKeywordsBlock` for getting related keywords
from DataForSEO Labs
- Implemented proper Pydantic models (`KeywordSuggestion` and
`RelatedKeyword`) for type-safe outputs
- Added mockable private methods (`_fetch_keyword_suggestions` and
`_fetch_related_keywords`) for better testability
- Included comprehensive test mocks to allow testing without actual API
credentials
- Both blocks support optional SERP info and clickstream data
- Added DataForSEO provider configuration using the SDK's
ProviderBuilder pattern
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
<!-- Put your test plan here: -->
- [x] Run block tests for DataForSeoKeywordSuggestionsBlock
- [x] Run block tests for DataForSeoRelatedKeywordsBlock
- [x] Verify mocks work correctly without API credentials
- [x] Confirm proper Pydantic model serialization
- [x] Run poetry format and fix any linting issues
This PR transforms GitHub Copilot from a general coding assistant into a
platform-specific expert by providing comprehensive onboarding and a
fully configured development environment that mirrors the AutoGPT
platform's CI/CD pipeline.
## Key Features
### CI-Aligned Development Environment
- **Production-Grade Setup**: The `copilot-setup-steps.yml` workflow now
mirrors the platform's CI workflows, providing Copilot with the same
environment used for testing and deployment
- **Essential Services**: Automatically starts Docker services
(PostgreSQL, Redis, RabbitMQ, ClamAV) required for platform development
- **Database Readiness**: Includes Prisma migrations and client
generation to ensure the database is immediately usable
- **Environment Configuration**: Copies default environment files and
validates service health, eliminating manual setup steps
### Comprehensive Platform Knowledge
- **Unified Documentation**: The `copilot-instructions.md` file
integrates knowledge from multiple sources (installer.md,
advanced_setup.md, CLAUDE.md, AGENTS.md) into a single comprehensive
guide
- **Development Patterns**: Includes advanced patterns for block
development, API creation, frontend work, and security implementation
- **Architecture Insights**: Provides deep understanding of the agent
block system, database schema, middleware, and CI/CD workflows
- **Troubleshooting Guide**: Comprehensive error handling and validation
steps based on production experience
### Enhanced Development Workflow
```bash
# Before: Trial-and-error dependency installation
# After: Fully configured environment ready immediately
# Copilot now understands:
poetry run serve # Backend server (port 8000)
pnpm dev # Frontend server (port 3000)
docker compose up -d # Essential services
poetry run pytest backend/blocks/test/test_block.py -xvs # Block validation
```
## Benefits
This configuration provides Copilot with:
- **Immediate Productivity**: No time wasted on environment setup or
dependency discovery
- **Platform Expertise**: Deep understanding of AutoGPT's architecture,
security patterns, and development workflows
- **CI Consistency**: Local development environment matches production
testing environment
- **Comprehensive Context**: Access to all platform documentation and
development patterns in a single source
The setup eliminates the slow trial-and-error process typical of
LLM-based development tools and ensures Copilot works efficiently from
the first interaction.
## References
- [GitHub Copilot Agent Environment
Setup](https://docs.github.com/en/copilot/how-tos/use-copilot-agents/coding-agent/customize-the-agent-environment#preinstalling-tools-or-dependencies-in-copilots-environment)
- [AutoGPT Platform CI Workflows](.github/workflows/)
- [Platform Development Guide](autogpt_platform/CLAUDE.md)
<!-- START COPILOT CODING AGENT TIPS -->
---
💡 You can make Copilot smarter by setting up custom instructions,
customizing its development environment and configuring Model Context
Protocol (MCP) servers. Learn more [Copilot coding agent
tips](https://gh.io/copilot-coding-agent-tips) in the docs.
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: ntindle <8845353+ntindle@users.noreply.github.com>
Co-authored-by: Nicholas Tindle <nicholas.tindle@agpt.co>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Nicholas Tindle <nicktindle@outlook.com>
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
- Resolves#10653
The objective is to move to a base image with fewer active
vulnerabilities. Hence the choice for `debian:13-slim` (0 high, 1
medium, 21 low severity), a huge improvement compared to our current
base image `python:3.11.10-slim-bookworm` (4 high, 11 medium, 15 low
severity).
### Changes 🏗️
- Change backend base image to `debian:13-slim`
- Use Python 3.13
- Fix now-deprecated use of class property in `AppProcess` and
`BaseAppService`
- Expand backend CI matrix to run with Python 3.11 through 3.13
- Update Python version constraint in `pyproject.toml` to include Python
3.13
Also, unrelated:
- Update `autogpt-platform-backend` package version to `v0.6.22`, the
latest release
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] CI passes
- [x] No new errors in deployment logs
- [x] Everything seems to work normally in deployment
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot merge` will merge this PR after your CI passes on it
- `@dependabot squash and merge` will squash and merge this PR after
your CI passes on it
- `@dependabot cancel merge` will cancel a previously requested merge
and block automerging
- `@dependabot reopen` will reopen this PR if it is closed
- `@dependabot close` will close this PR and stop Dependabot recreating
it. You can achieve the same result by closing it manually
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore <dependency name> major version` will close this
group update PR and stop Dependabot creating any more for the specific
dependency's major version (unless you unignore this specific
dependency's major version or upgrade to it yourself)
- `@dependabot ignore <dependency name> minor version` will close this
group update PR and stop Dependabot creating any more for the specific
dependency's minor version (unless you unignore this specific
dependency's minor version or upgrade to it yourself)
- `@dependabot ignore <dependency name>` will close this group update PR
and stop Dependabot creating any more for the specific dependency
(unless you unignore this specific dependency or upgrade to it yourself)
- `@dependabot unignore <dependency name>` will remove all of the ignore
conditions of the specified dependency
- `@dependabot unignore <dependency name> <ignore condition>` will
remove the ignore condition of the specified dependency and ignore
conditions
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Swifty <craigswift13@gmail.com>
With this PR, when a user removes the agent file, the agent name and
description will also be removed if they haven’t been changed. However,
if the user has modified either of them, they will remain.
### Checklist 📋
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Everything works perfectly on local, and the tests are also
passing.
- resolves -
https://github.com/Significant-Gravitas/AutoGPT/issues/10511
In this PR, I’ve added backend endpoints and a frontend UI for edit
functionality on the Agent Dashboard. Now, users can update their store
submission, if status is `PENDING` or `APPROVED`, but not for `REJECTED`
and `DRAFT`. When users make changes to a pending status submission, the
changes are made to the same version. However, when users make changes
to an approved status submission, a new store listing version is
created.
Backend works something like this:
<img width="866" height="832" alt="Screenshot 2025-08-15 at 9 39 02 AM"
src="https://github.com/user-attachments/assets/209c60ac-8350-43c1-ba4c-7378d95ecba7"
/>
### Changes
- I’ve updated the `StoreSubmission` view to include `video_url` and
`categories`.
- I’ve added a new frontend UI for editing submissions.
- I’ve created an endpoint for editing submissions.
- I’ve added more end-to-end tests to ensure the edit submission
functionality works as expected.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] I have checked manually, everything is working perfectly.
- [x] All e2e tests are also passing.
---------
Co-authored-by: Zamil Majdy <zamil.majdy@agpt.co>
Co-authored-by: neo <neo.dowithless@gmail.com>
Co-authored-by: Reinier van der Leer <pwuts@agpt.co>
Co-authored-by: Swifty <craigswift13@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Ubbe <hi@ubbe.dev>
Co-authored-by: Lluis Agusti <hi@llu.lu>
Some API key endpoints have the Launch Darkly feature flag enabled,
while others don’t. To ensure consistency and remove the API key flag
from the Launch Darkly dashboard, I’m also removing it from the left
endpoints.
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Everything is working fine locally
I’ve added three tests for the API keys page:
- The test checks if the user is redirected to the login page when
they’re not authenticated.
- The test verifies that a new API key is created successfully.
- The test ensures that an existing API key can be revoked.
<img width="470" height="143" alt="Screenshot 2025-08-19 at 10 56 19 AM"
src="https://github.com/user-attachments/assets/d27bf736-61ec-435b-a6c4-820e4f3a5e2f"
/>
I’ve also removed the feature flag from the `delete_api_key` endpoint,
so we can use it on CI and in the local environment.
### Checklist 📋
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] tests are working perfectly locally.
---------
Co-authored-by: Ubbe <hi@ubbe.dev>
## Changes 🏗️
Setup for the new Agent Runs page:
<img width="900" height="521" alt="Screenshot 2025-08-15 at 14 36 34"
src="https://github.com/user-attachments/assets/460d6611-4b15-4878-92d3-b477dc4453a9"
/>
It is behind a feature flag in Launch Darkly, `new-agent-runs`, so we
can progressively enable in staging and later on production.
### Other improvements
<img width="350" height="291" alt="Screenshot_2025-08-15_at_14 28 08"
src="https://github.com/user-attachments/assets/972d2a1a-a4cd-4e92-b6d7-2dcf7f57c2db"
/>
- Added a new `<ErrorCard />` component to paint gracefully API errors
when fetching data
- Moved some sub-components of the old library page to a nested
`/components` folder 📁
Behind a feature flag
## Checklist 📋
### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Tested with the feature flag ON and OFF
### For configuration changes:
None
---------
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
In this project, I’ve added all the reusable, non-reactive components
that will be used in the new block menu. I’ve also included a new
library called `react-timeago` that helps us find related times.
### Checklist 📋
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Everything works perfectly locally
## Changes 🏗️
- Run the API query generation as part of the `dev` command
- update the `README` to reflect so
- Add CI job to generate queries and type-check to make sure we are not
out of sync
- the job is run both in Front-end and Back-end changes
- Generate the files via script to load the BE URL dynamically from the
env
- Remove generated files from Git
- rename the `type-check` command to `types`
## Checklist 📋
### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] CI passes
- [x] `README` updates make sense
#### For configuration changes:
None
---------
Co-authored-by: Zamil Majdy <zamil.majdy@agpt.co>
## Changes 🏗️
We had 2 flaky end-to-end tests:
- Build page → user can add two blocks and connect them
- this was failing sometimes because the `Run` button on the builder
does not work well, sometimes you need to click it twice for it to
work...
- Agent dashboard → edit actions
- some flaky tests asserting agent submissions not being there, pulled
the fixes from Abhi here on this PR
https://github.com/Significant-Gravitas/AutoGPT/pull/10545
## Checklist 📋
### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] E2E pass on the CI
- [x] Changes make sense
### For configuration changes:
None
## Changes 🏗️
Add the following caps to the **Agent Activity Dropdown**:
- display activity only from the last 72h
- display up to 1000 items
## Checklist 📋
### For code changes
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Open the agent activity with a big amount of times locally
- [x] It displays up to a 1000 and with 72h cap
### For configuration changes
None
Co-authored-by: Abhimanyu Yadav <122007096+Abhi1992002@users.noreply.github.com>
When the FirecrawlExtractBlock receives an output_schema, we currently
declare the field as a str.
Pydantic therefore serialises the JSON‐looking value into a string and
the Firecrawl API rejects the request with:
`400 Bad Request – Invalid JSON schema. path: ['schema']`
Direct curl requests work because the same structure is sent as a proper
JSON object.
### Changes 🏗️
- Changed the output_schema to dict instead of str
### Checklist 📋
#### For code changes:
- [x] I have clearly listed my changes in the PR description
- [x] I have made a test plan
- [x] I have tested my changes according to the test plan:
- [x] Test firebase.extract(..., schema) works with dict rather than str
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
