Change TS client search() return type from Promise<Array<Memory>>
to Promise<{ results: Array<Memory> }> to match v3 API envelope.
Update unit test mocks to return {results: []} and fix assertions.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Remove test_neo4j_cypher_syntax.py and test_graph_delete_docker.py
which reference deleted graph_memory module.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Strip graph_memory patches, GraphStoreFactory mocks, graph_store config,
enable_graph parametrize, _add_to_graph assertions, and "relations"
assertions from all test fixtures and test functions. Graph store has
been removed from the codebase.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
v3 search returns {results: [...]} instead of a bare array.
Updated waitForSearchResults helper and search edge case tests
to unwrap the v3 envelope.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Restore LMStudioEmbedder, LMStudioLLM, DeepSeekLLM in TS factory.ts
(accidentally removed during graph store cleanup — unrelated to graphs)
- Add langchain>=0.1.0 to Python extras dep group (was transitively
provided by langchain-neo4j in the removed graph group, still needed
by mem0/llms/langchain.py and mem0/embeddings/langchain.py)
- Add pytest.importorskip("langchain") to test_langchain.py for graceful
skip when langchain isn't installed
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
These test files import mem0.memory.graph_memory, kuzu_memory,
memgraph_memory, apache_age_memory, and mem0.graphs.neptune which
were deleted in the graph store removal commit.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Update mock URLs in memoryClient.crud.test.ts and
memoryClient.search.test.ts to match v3 endpoints:
- add: /v1/memories/ -> /v3/memories/
- search: /v2/memories/search/ -> /v3/memories/search/
- deleteAll: stays on /v1/memories/ (no v3 equivalent)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Was missing 13 entries (credentials, token, access_token, refresh_token,
auth_token, session_token, client_secret, auth_client_secret,
azure_client_secret, service_account_json, aws_session_token, credential,
secret). Now matches the original list from commit 46b4b2e9.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The function and associated constants (_SENSITIVE_FIELDS_EXACT,
_SENSITIVE_SUFFIXES) were accidentally dropped during the merge.
Restores telemetry-safe config cloning that redacts API keys,
passwords, and other secrets while preserving runtime auth objects.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Remove deprecated custom_update_memory_prompt from MemoryConfig entirely
- Add MAX_BATCH=100 chunking guard to embed_batch in openai.py and azure_openai.py
- Move all inline imports (mem0.utils.*) to top-level in main.py
- Remove custom_update_memory_prompt from test fixture
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Avoids spawning 2 threads per add() call when graph is not configured.
Fixes thread leakage under high concurrency (was accumulating 3500+ threads
per worker). Calls _add_to_vector_store directly in the current thread.
Applied to both sync Memory and async AsyncMemory classes.
- Replace sequential _upsert_entity() loop with batched pipeline:
1. Global dedup: collect unique entities across all memories
2. Single embed_batch() call for all entity texts
3. Batch search via search_batch() (Qdrant query_batch_points)
4. Batch insert for new entities, individual updates for existing
- Add search_batch() to VectorStoreBase (sequential fallback)
- Add native search_batch() to Qdrant using query_batch_points
- ~6-10x reduction in entity linking network round trips
Streamline prompt guidelines and update examples for clarity.
Remove redundant sections and verbose explanations.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Move spacy from core dependencies to optional [nlp] extra.
The system gracefully degrades without spaCy:
- Lemmatization falls back to raw text
- Entity extraction returns empty (no entity boost)
- BM25 keyword search still works on raw text
- Semantic search and additive scoring unaffected
Install with NLP support: pip install mem0ai[nlp]
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add text_lemmatized to _create_memory() and _update_memory() (sync + async)
so BM25/keyword search works for all code paths, not just the batch pipeline
- pgvector: add GIN index on text_lemmatized for fast full-text search
- pgvector: add error logging to keyword_search
- milvus: add sparse BM25 field + function to create_col() so keyword_search works
- mongodb: auto-create Atlas Search text index in create_col()
- azure_mysql: add generated column + FULLTEXT index for keyword_search
- Add sparse_vectors_config with 'bm25' named vector to create_col()
- Compute BM25 sparse vectors from text_lemmatized during insert/update
- Use client-side fastembed Qdrant/bm25 encoder instead of server-side inference
- Graceful fallback if fastembed not installed
- Fixes keyword_search() returning None on self-hosted Qdrant
- Rename `observation_date` param to `timestamp` (matches platform)
- Change `rerank` default from True to False (matches platform)
- Internal prompt builder still uses "Observation Date" section heading
for LLM compatibility with ADDITIVE_EXTRACTION_PROMPT
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace the 2-LLM-call add pipeline with a single-pass additive extraction
using ADDITIVE_EXTRACTION_PROMPT. Memories now accumulate (ADD-only) instead
of being updated/deleted during extraction.
Search pipeline upgraded to hybrid scoring combining three signals:
- Semantic similarity (vector search)
- BM25 keyword matching (native per vector store, 15 stores supported)
- Entity boost (spaCy NER with entity collection linking)
Combined score = (semantic + bm25 + entity_boost) / max_possible, where
max_possible adapts based on which signals are active.
Key changes:
- Add spaCy-based lemmatization for BM25 keyword search
- Add spaCy-based entity extraction (PROPER, QUOTED, COMPOUND, NOUN types)
- Add entity store as second vector collection ({collection}_entities)
- Add native keyword_search() to 15 vector store adapters
- Add batch embedding support (embed_batch) for OpenAI and Azure OpenAI
- Add message persistence in SQLite (rolling window of 10 per scope)
- Add additive scoring with adaptive normalization
- Add observation_date parameter to add()
- Add custom_instructions config field
- Default search threshold changed to 0.1
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
