feat(memory): capture→distill v2 and Memento client integration - #169
Merged
Merged
Conversation
- Introduced a new observation layer for capturing user insights without early classification. - Updated the memory extraction workflow to persist observations and trigger distillation. - Refactored related services and repository methods to support the new observation model. - Added tests for the new capture prompt and ensured existing functionality remains intact. This change enhances the ability to gather user concerns and insights for future processing.
…agent harness - Added a new memory search tool to retrieve insights from fluid memory, allowing for more context-aware responses. - Updated the agent harness to utilize memory tools, ensuring that memory searches are prioritized before document retrieval. - Introduced memory citations in output manifests to track references from memory searches. - Enhanced tests to validate the new memory search behavior and its integration with existing functionalities. This update improves the agent's ability to leverage past interactions for more relevant and informed responses.
- Updated `.env.local.example` to reflect the new Knowhere API base URL and added Memento service configuration. - Added `@modelcontextprotocol/sdk` dependency to `package.json` and updated `pnpm-lock.yaml` accordingly. - Removed deprecated constants and scripts related to the 观心 case, streamlining the codebase. - Refactored agent harness to utilize the new memory capture functionality, enhancing memory management during chat interactions. - Updated tests to reflect changes in memory handling and ensure proper integration with the new SDK. This update improves the overall functionality and maintainability of the memory and chat systems.
… configuration - Changed the namespace from "notebook-namespace" to "default" in chat service tests to align with new defaults. - Updated the chat service to use a shared library namespace instead of fetching compatible namespaces. - Increased the timeout for the Knowhere client to 120 seconds to enhance reliability during API calls. These changes streamline namespace management and improve the integration with the Knowhere API.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
1 Skipped Deployment
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
fluid_observations) instead of single-pass direct write.commitChatTurnorchestration for post-answer memory capture and activation recording.Test plan
vitest run— 803 passed, 14 skippedNotes
Distill batch trigger (pending ≥ 8) is implemented in this branch for the in-notebook path. Follow-up Memento service decoupling is tracked separately in Ontos-AI/Memento.
Made with Cursor