Remove the deprecated expired_at field from all graph models, Neo4j
Cypher queries, repositories, and pipeline code. Replace with dialog_at
on StatementNode to track the original dialog timestamp.
- Strip expired_at from DialogueNode, ChunkNode, StatementNode,
ExtractedEntityNode, edges, and all Cypher queries
- Add dialog_at to MessageItem schema and propagate through extraction
and graph build steps
- Extract emotion/metadata async submission from WritePipeline into
a generic _submit_celery_task helper
- Add post_store_dedup_and_alias_merge Celery task for async alias
merging and second-layer dedup after Neo4j write
- Switch pytest async backend from anyio to asyncio_mode=auto
- Replace extract_user_metadata_task with entity-level extract_metadata_batch_task
- Add MetadataExtractionStep following ExtractionStep pattern with Jinja2 prompts
- Flatten MetadataExtractionResponse to 9-field schema (aliases, core_facts, etc.)
- Add Cypher queries for incremental metadata writeback and alias edge redirection
- Wire _extract_metadata into WritePipeline as Step 3.6 (fire-and-forget)
- Add pilot_write() to MemoryService; refactor pilot_run_service to use it
- Extract snapshot logic into WriteSnapshotRecorder
- Add valid_at/invalid_at passthrough in triplet extraction prompt (both zh/en)
- Propagate temporal_validity to EntityEntityEdge in ExtractionOrchestrator
- Use coalesce() for valid_at/invalid_at in Neo4j cypher queries to handle NULLs
- Fix workspace_id/config_id UUID parsing in read_memory config resolution
- Downgrade verbose extraction pipeline logs from info to debug
- Remove UUID and short API key patterns from sensitive filter to reduce false positives
- Standardize log message format (use = spacing, end_user_id label)
- Fix misindented TODO comment in write_pipeline.py
- Rename StatementExtractionStep → StatementTemporalExtractionStep and
extract_statement.jinja2 → extract_statement_temporal.jinja2 to reflect
merged temporal extraction logic
- Move extraction_pipeline_orchestrator.py out of steps/ to engine root
- Move dedup_step.py into steps/ directory
- Introduce WriteMemoryRequest schema to replace positional args in write_memory()
- Extract _resolve_and_load_config, _preprocess_files, _write_neo4j, and
_invalidate_interest_cache as private helpers in MemoryAgentService
- Remove shadow pipeline and simplify NEW_PIPELINE_ENABLED branch
- Merge 类型归属/成员隶属/任职服务 relation types into single 归属身份关系 in triplet prompt
- Add alias merge logic (别名属于) in deduplication and MERGE_ALIAS_BELONGS_TO Cypher query
- Add StorageType, Language, MessageItem enums/models to memory_agent_schema
- Reduce AgentMemory_Long_Term.DEFAULT_SCOPE from 6 to 1
- Delete standalone extract_temporal.jinja2 (logic merged into statement step)
Introduce a layered pipeline architecture for the memory write flow:
- WritePipeline: orchestrates preprocess → extract → store → cluster → summarize
with deadlock retry, resource cleanup, and pilot-run support
- MemoryService: facade that delegates to WritePipeline, placeholder methods
for read/forget/reflect
- BearLogger: structured step-level logging with perf threshold alerts
- Shadow pipeline integration in MemoryAgentService (env-gated pilot run)
Also includes:
- Fix deprecated SQLAlchemy declarative_base import
- Extend Neo4j Entity fulltext index to cover description and aliases
- Migrate Pydantic schemas to v2 (ConfigDict, field_validator)
- Consolidate memory search services by removing separate content_search.py and perceptual_search.py
- Update model client handling in base_pipeline.py to use ModelApiKeyService for LLM client initialization
- Add new prompt files and modify existing services to support consolidated search architecture
- Refactor memory read pipeline and related services to use updated model client approach
- Consolidate memory search services by removing separate content_search.py and perceptual_search.py
- Update model client handling in base_pipeline.py to use ModelApiKeyService for LLM client initialization
- Add new prompt files and modify existing services to support consolidated search architecture
- Refactor memory read pipeline and related services to use updated model client approach
- Skip alias merging for user entities during dedup (_merge_attribute and
_merge_entities_with_aliases) to prevent dirty data from overwriting
PgSQL authoritative aliases
- Add PgSQL→Neo4j alias sync after Neo4j write in write_tools to
ensure Neo4j user entities always reflect the PgSQL source
- Remove deduped_aliases (Neo4j history) from alias sync in
extraction_orchestrator, only append newly extracted aliases to PgSQL
- Guard Neo4j MERGE cypher to preserve existing aliases for user
entities (name IN ['用户','我','User','I'])
- Fix emotion_analytics_service query to use ExtractedEntity label
and entity_type property
- Replace storage_services/search with new read_services/memory_search structure
- Implement content_search and perceptual_search strategies
- Add query_preprocessor for search optimization
- Create memory_service as unified interface
- Update celery_app and graph_search for new architecture
- Add enums for memory operations
- Implement base_pipeline and memory_read pipeline patterns
- Merge alias add/remove into MetadataExtractionResponse and Celery metadata task,
removing the separate sync step from extraction_orchestrator
- Replace first-person pronouns ("我") with "用户" in statement extraction to
preserve identity semantics for downstream metadata/alias extraction
- Update extract_statement.jinja2 prompt to enforce "用户" as subject for user
statements instead of resolving to real names
- Add alias change instructions (aliases_to_add/aliases_to_remove) to
extract_user_metadata.jinja2 with incremental merge logic
- Deduplicate special entities ("用户", "AI助手") in graph_saver by reusing
existing Neo4j node IDs per end_user_id
- Sync final aliases from PgSQL to Neo4j user entity nodes after metadata write
- Adjust multi-modal memory write behavior for text and visual data
- Mask API keys in model list response to prevent exposure
- Add capability-based filtering to the model list API
* [fix]The log retains genuine alerts and errors, while filtering out unnecessary noise.
* [fix]Scenario English and Chinese, emotion specifications
* [fix]Change the "no data" scenario from 0.0 to None
* [fix]The emotional health indicators, emotional advice, and emotional distribution analysis are all linked together.
* [fix]The emotional health indicators, emotional advice, and emotional distribution analysis are all linked together.
* [fix]Separate expected errors from unexpected errors
