-
Notifications
You must be signed in to change notification settings - Fork 12
Add Mulit-agent Implementation with Web Search Tool #38
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
base: main
Are you sure you want to change the base?
Conversation
- Updated .env.example to include WEB_SEARCH_BASE_URL and WEB_SEARCH_API_KEY. - Introduced web_search.py to implement a tool for fetching Google Search grounded responses from the Gemini model.
…h query extraction - Introduced RETRYABLE_STATUS for managing retryable HTTP status codes. - Updated the _post method to _post_payload for clarity and added a docstring. - Enhanced error handling in the backoff decorator to include specific exceptions and retry logic based on RETRYABLE_STATUS.
…ValueError for missing environment variables.
- Updated the add_citations function to return a tuple of synthesized text and a mapping of citation IDs to source labels. - Enhanced the GroundedResponse class to include a citations field. - Improved error handling and type checks throughout the citation processing logic. - Introduced a new _collect_citations function to streamline citation ID collection from candidates.
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 adds a multi-agent implementation that combines knowledge base search with web search capabilities through Gemini's grounding feature. The implementation uses a planner-worker architecture where a main agent coordinates between a knowledge base agent and a web search tool to answer complex queries that may require both historical and current information.
Key Changes:
- Implements
GeminiGroundingWithGoogleSearchtool for web search with automatic citation generation - Creates a multi-agent system with dedicated knowledge base and web search capabilities
- Adds test coverage for the web search integration
Reviewed Changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
src/utils/tools/gemini_grounding.py |
Implements the Gemini grounding tool with Google Search integration, including citation parsing and error handling |
src/2_frameworks/2_multi_agent/efficient_multiple_kbs.py |
Creates a planner-worker multi-agent system that routes queries between knowledge base and web search tools |
tests/tool_tests/test_gemini_grounding.py |
Adds integration test for the web search grounding functionality |
.env.example |
Adds required environment variables for web search API configuration |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| try: | ||
| response.raise_for_status() | ||
| except httpx.HTTPStatusError as exc: | ||
| raise exc from exc |
Copilot
AI
Nov 5, 2025
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.
Re-raising the exception with itself as the cause (raise exc from exc) is redundant. Either raise without the from clause or re-raise as a different exception type with the original as the cause.
| raise exc from exc | |
| raise exc |
PR Type
Feature
Short Description
Tests Added
...