-
Notifications
You must be signed in to change notification settings - Fork 1
feature: moved the agentic layer from a standalone service to a lightweight client example, Kling / seedance minor version updates #66
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
…getting replicate results. Improved logging
… not really a full feature of the stack
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Pull request overview
This PR refactors the agentic layer from a standalone service to a lightweight client example, updates model versions, improves worker logging, and migrates from the requests library to httpx for better async support.
Changes:
- Removed the agentic Docker service and restructured it as a client example under
clients/agentic/ - Updated Kling video model from v2.5-turbo-pro to v2.6 and added audio generation control
- Migrated video download functionality from requests to httpx library
- Enhanced replicate polling with configurable intervals and better logging
Reviewed changes
Copilot reviewed 30 out of 31 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
| workers/videos/external/kling_2.py | Updated Kling model version to 2.6 and added generate_audio flag |
| workers/videos/context.py | Migrated from requests to httpx for video downloads with redirect support |
| workers/common/replicate_helpers.py | Added configurable poll_interval parameter and completion logging |
| workers/common/logger.py | Removed httpx/huggingface_hub logging suppression |
| scripts/package_release.py | Removed agentic service from release pinning |
| docker-compose.yml | Removed agentic service definition |
| deferred-diffusion.code-workspace | Removed agentic workspace folder |
| clients/agentic/requirements.txt | Added new agentic client dependencies |
| clients/agentic/main.py | Created new agentic client entry point with uvicorn |
| clients/agentic/agents/fetch_agent.py | Moved fetch agent to clients with updated imports |
| clients/agentic/agents/dd_mcp_server.py | Moved and simplified MCP server configuration |
| clients/agentic/agents/chat_agent.py | Moved chat agent to clients with updated imports |
| clients/README.md | Added documentation for agentic layer as client example |
| agentic/* | Removed old agentic service files (app.py, agents, views, etc.) |
| Dockerfile.agentic | Removed agentic Docker configuration |
| .github/workflows/release.yml | Removed agentic image build and tagging from CI/CD |
| .github/copilot-instructions.md | Updated documentation removing agentic service references |
| README.md | Removed agentic service documentation section |
Comments suppressed due to low confidence (1)
clients/agentic/agents/fetch_agent.py:9
- The model identifier "gpt-4.1-mini" does not appear to be a valid OpenAI model name. As of January 2025, OpenAI's model naming convention includes models like "gpt-4o-mini", "gpt-4-turbo", "gpt-4", etc., but not "gpt-4.1-mini". This should be verified against OpenAI's current model catalog to ensure the agent can actually connect to a valid model.
No description provided.