fleet-memory/hindsight-api/hindsight_api
Nicolò Boschi 5d05962db0
feat: add source facts token limits to consolidation and recall (#509)
* feat: add source facts token limits to consolidation and recall

- Add two new configurable (per-bank) parameters:
  - consolidation_source_facts_max_tokens: total token budget for source
    facts across all observations in the consolidation prompt (-1 = unlimited)
  - consolidation_source_facts_max_tokens_per_observation: per-observation
    cap so each observation gets a fair share of source facts (-1 = unlimited,
    default 256)
- Both are also exposed as recall API parameters via SourceFactsIncludeOptions
  (max_tokens and max_tokens_per_observation)
- Consolidation now uses resolve_full_config to respect bank-level overrides
- Improve consolidation prompt: temporal metadata (occurred_start=, | Involving:)
  is now clearly separated from observation text, with a concrete example showing
  the expected synthesis style and explicit rules not to copy raw fact lines
- Add tests for recall source facts capping and consolidation config forwarding
- Expose all three new fields in the control plane bank config UI
- Document new env vars in configuration.md
- Regenerate OpenAPI spec and all SDK clients

* fix: reorder observations UI fields and rename Label Groups to Entity Labels

* fix: revert Entities section title (only rename inner label)

* doc: add consolidation source facts and batch size fields to memory-banks docs
2026-03-06 13:01:33 +01:00
..
admin feat: new 'worker' service (#176) 2026-01-20 10:17:56 +01:00
alembic perf: add GIN index on source_memory_ids for observation lookup (#485) 2026-03-05 10:15:40 +01:00
api feat: add source facts token limits to consolidation and recall (#509) 2026-03-06 13:01:33 +01:00
engine feat: add source facts token limits to consolidation and recall (#509) 2026-03-06 13:01:33 +01:00
extensions Add on_file_convert_complete extension hook after file-to-markdown conversion (#507) 2026-03-06 09:56:53 +01:00
webhooks fix: use correct schema name in webhook outbox callback to prevent silent transaction rollback (#499) 2026-03-05 17:07:48 +01:00
worker feat: webhook system with retain.completed event, UI, and docs (#487) 2026-03-04 14:17:01 +01:00
__init__.py Release v0.4.16 2026-03-05 17:54:59 +01:00
banner.py feat: support for pgvectorscale (DiskANN) (#378) 2026-02-16 14:19:56 +01:00
config.py feat: add source facts token limits to consolidation and recall (#509) 2026-03-06 13:01:33 +01:00
config_resolver.py Fix bank config API for multi-tenant schema isolation (#417) 2026-02-20 23:43:52 +01:00
daemon.py feat: improve openclaw and hindisght-embed params (#279) 2026-02-03 09:39:04 +01:00
main.py feat: add source facts token limits to consolidation and recall (#509) 2026-03-06 13:01:33 +01:00
mcp_local.py fix(mcp): unify hindsight-mcp-local and server mcp (#407) 2026-02-19 17:57:17 +01:00
mcp_tools.py Fix bank-level MCP tool filtering for FastMCP 3.x (#491) 2026-03-04 10:29:29 -05:00
metrics.py chore: remove dead code (#245) 2026-01-30 09:16:32 +01:00
migrations.py fix(performance): improve recall and retain performance on large banks (#469) 2026-03-03 13:35:22 +01:00
models.py Add bank-scoped validation to engine and HTTP handlers (#454) 2026-03-02 09:55:21 +01:00
pg0.py feat: support vertex as llm provider (#233) 2026-01-29 16:13:57 -05:00
server.py Fix: Load extensions in server.py for multi-worker deployments (#155) 2026-01-13 17:55:33 +01:00
tracing.py feat: add otel traceability (#330) 2026-02-10 12:20:48 +01:00
utils.py fix(helm): improve appVersion usage (#326) 2026-02-09 11:35:08 +01:00