Fixed #62 : unify AI agents under FastMCP and implement hybrid RAG#73
Fixed #62 : unify AI agents under FastMCP and implement hybrid RAG#73debanganghosh08 wants to merge 5 commits intoopenMF:mainfrom
Conversation
There was a problem hiding this comment.
Why are you moving the imports inside the function?
There was a problem hiding this comment.
You're right, thank you for pointing that out! I've moved the imports back to the top level in the latest commit to follow standard Python conventions. I had originally moved them to troubleshoot a local environment conflict, but that is no longer necessary.
There was a problem hiding this comment.
We aren't really using this so no need to make any changes here, it's mostly legacy code and I am unable to understand what are you trying to change?
There was a problem hiding this comment.
Understood! I've reverted all changes to repo_cloner.py to keep this PR strictly focused on the FastMCP unification and avoid adding noise to legacy code. Thanks for clarifying its status.
2 participants
I have implemented a unified AI agent architecture to resolve the fragmentation described in Issue #62. By migrating standalone scripts to a central FastMCP server, the system now supports real-time technical reasoning across multiple data sources.
Key Technical Implementations -
✅ Unified FastMCP Server: Created a central hub in MCP_Enhancement/agent_start.py that exposes Jira, GitHub, and Documentation search as discoverable tools for the AI.
✅ Jira Comment Extraction: Upgraded the Jira tool to fetch the last 3 developer comments per ticket, ensuring the AI captures the "consensus" and technical decisions, not just the status.
✅ Cross-System Synthesis: Added a generate_knowledge_summary tool that automatically links Jira keys to their respective GitHub Pull Requests to provide a holistic view of feature development.
✅ Hybrid RAG Integration: Implemented a search_project_docs tool that allows the agent to search through static Markdown files while maintaining live connections to operational APIs.
Why this is better -
✅ Scalability: New tools (like Slack or Ledger logs) can now be added to the same server with just a few lines of code.
✅ Maintainability: Eliminates duplicate boilerplate code (LLM initialization, env loading) found in the previous standalone scripts.
✅ Real-time Accuracy: Replaces brittle, snapshot-based RAG with live API queries via the Model Context Protocol.
Thank you so much for reading, Looking forward for your review and Approval. 😇