Add a long-term memory interface #25
Open
Conversation
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
- Removed duplicate session endpoints from api.py in favor of v1 API structure - Kept vectorstore adapter approach in long_term_memory.py over direct Redis operations - Used feature branch version of uv.lock for latest dependencies 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <[email protected]>
- Configure agent-memory-client as workspace member - Fix ULID field generation in LenientMemoryRecord model - Note: Some tests need updating for new MemoryRecord API structure 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <[email protected]>
There was a problem hiding this comment.
Pull Request Overview
Introduce a pluggable VectorStoreAdapter interface to support multiple LangChain vector backends for long-term memory storage, with Redis as the default.
- Define
VectorStoreAdapterabstraction withLangChainVectorStoreAdapterandRedisVectorStoreAdapter. - Refactor core long‐term memory functions (
index_,search_,count_) to delegate to the adapter. - Update configuration, dependencies, documentation, and tests to cover new backends.
Reviewed Changes
Copilot reviewed 20 out of 22 changed files in this pull request and generated 1 comment.
Show a summary per file
| File | Description |
|---|---|
| agent_memory_server/vectorstore_adapter.py | Added adapter interface and implementations |
| agent_memory_server/vectorstore_factory.py | Factory logic for selecting and creating adapters |
| agent_memory_server/long_term_memory.py | Refactored indexing/search/count to use adapter |
| agent_memory_server/config.py | Extended settings for backend and RedisVL names |
| pyproject.toml | Updated dependencies and extras for backends |
| docs/vector-store-backends.md | New documentation listing supported backends |
| tests/test_vectorstore_adapter.py | Added unit tests for adapter methods |
| tests/test_memory_compaction.py | Updated compaction tests to use adapter |
| tests/test_long_term_memory.py | Refactored long-term memory tests to patch adapter |
| tests/conftest.py | Patching of Redis connections and global state |
Comments suppressed due to low confidence (3)
tests/test_memory_compaction.py:165
- [nitpick] Remove debug print statements to keep test output clean; use assertions or logs if inspection is needed.
print(f"🔍 Redis keys after indexing: {keys}")
tests/test_vectorstore_adapter.py:221
- Add a test case for
delete_memorieswith non-empty input to verify that deletion logic returns the correct count and propagates to the underlying store.
deleted = await adapter.delete_memories([])
tests/test_memory_compaction.py:176
- [nitpick] Remove debug print statements to keep test output clean; rely on assertions rather than printing.
)
Co-authored-by: Copilot <[email protected]>
|
@claude Resolve these conflicts. |
|
Claude encountered an error —— View job I'll analyze this and get back to you. |
- Resolved ULID import conflicts in models and client files - Fixed README.md merge conflict preserving vector store backends doc - Resolved pyproject.toml dependency conflicts - Fixed test file conflicts and cleaned up malformed merge resolutions - Preserved all vector store adapter functionality and interface - Tests passing, vector store factory working correctly 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Move vector search over long-term memories into a new
VectorStoreAdapterinterface instead of only including a concrete Redis implementation. Satisfies the new interface with a generic adapter supporting LangChainVectorStoreinstances. Then satisfies the same interface with the langchain-redisRedisVectorStore, making Redis the default long-term storage layer.