# Flumes AI

## Docs

- [Build magical context for an agent turn (optionally extract & ingest).](https://docs.flumes.ai/api-reference/context/build-magical-context-for-an-agent-turn-optionally-extract-&-ingest.md)
- [List memories with filters (paginate).](https://docs.flumes.ai/api-reference/memories/list-memories-with-filters-paginate.md)
- [Soft-delete (archive) a memory (hard delete may come later).](https://docs.flumes.ai/api-reference/memories/soft-delete-archive-a-memory-hard-delete-may-come-later.md)
- [Store a memory (fact, event, document, note, profile, chunk).](https://docs.flumes.ai/api-reference/memories/store-a-memory-fact-event-document-note-profile-chunk.md)
- [Update selected fields of a memory.](https://docs.flumes.ai/api-reference/memories/update-selected-fields-of-a-memory.md)
- [Hybrid retrieval over facts/events/documents with tunable weights.](https://docs.flumes.ai/api-reference/retrieval/hybrid-retrieval-over-factseventsdocuments-with-tunable-weights.md)
- [API metadata & limits.](https://docs.flumes.ai/api-reference/system/api-metadata-&-limits.md)
- [Liveness & readiness.](https://docs.flumes.ai/api-reference/system/liveness-&-readiness.md)
- [Authentication & Security](https://docs.flumes.ai/guides/concepts/authentication-security.md): Keep your data – and your users – safe.
- [Conflict Keys & Supersession](https://docs.flumes.ai/guides/concepts/conflict-keys-supersession.md): Keep one current truth per fact and audit prior values.
- [Entities & Identity](https://docs.flumes.ai/guides/concepts/entities-and-identity.md): Mint stable IDs, alias/merge, and scope data correctly.
- [Memory Types](https://docs.flumes.ai/guides/concepts/memory-types.md): Flumes automatically classifies each memory so you can filter and reason about them easily.
- [Observability](https://docs.flumes.ai/guides/concepts/observability.md): Events, request IDs, and inspecting costs/latencies.
- [PII Redaction & Sensitivity](https://docs.flumes.ai/guides/concepts/pii-redaction.md): Control masking at assemble time and per-memory sensitivity.
- [Retrieval Presets](https://docs.flumes.ai/guides/concepts/retrieval-presets.md): Balanced, factual, recent, and graphy — what they tune and when to use them.
- [Token Budgeting](https://docs.flumes.ai/guides/concepts/token-budgeting.md): How max_context_tokens works and how to interpret budget_actions.
- [Quickstart](https://docs.flumes.ai/guides/getting-started/quickstart.md): Store and retrieve your first memory in less than 5 minutes.
- [Integrating with an LLM Agent](https://docs.flumes.ai/guides/integration/integrating-llm-agent.md): Give your LangChain or custom agent persistent, scoped memory.
- [Memory Model Explained](https://docs.flumes.ai/guides/overview/memory-model.md): The schema we guarantee, how supersession works, and how retrieval blends signals.
- [Error Codes & Troubleshooting](https://docs.flumes.ai/guides/troubleshooting/error-troubleshooting.md): Resolve issues fast and keep shipping.
- [Flumes: Memory & Context API](https://docs.flumes.ai/index.md): Ingest facts/events, assemble context in one call, and get structured facts, a compact summary, and sources.
- [SDK Overview](https://docs.flumes.ai/sdk/index.md): Flumes SDKs – entity-first clients, simple add/search helpers, and an optional Agent helper with explicit LLM keys.
- [Python SDK](https://docs.flumes.ai/sdk/python.md): How to use the Flumes Python SDK – add, search, retrieval knobs, budgeting, and the optional Agent helper.

## OpenAPI Specs

- [flumes](https://docs.flumes.ai/api-reference/flumes.yaml)