atomicmemory-docs

Documentation site for AtomicMemory — the standardized platform layer for AI memory.

Live site: docs.atomicmemory.ai

Built with Docusaurus. Deployed to GitHub Pages.

Structure

docs/
├─ introduction.md          # Landing page (served at /)
├─ quickstart.md            # Docker → first ingest/search in 2 minutes
├─ platform/                # Why modular — the differentiator narrative
│  ├─ architecture.md       # Ingest / Search / CRUD / Lifecycle / Trust separation
│  ├─ composition.md        # createCoreRuntime, createApp, bindEphemeral seams
│  ├─ stores.md             # MemoryStore / SearchStore / ClaimStore / EntityStore / EpisodeStore
│  ├─ providers.md          # Pluggable embedding + LLM
│  ├─ scope.md              # MemoryScope (user / workspace / agent)
│  └─ observability.md      # RetrievalResult.observability, trace schemas
└─ api-reference/
   ├─ http/                 # Generated from @atomicmemory/atomicmemory-core's OpenAPI spec
   └─ sdk/                  # SDK method reference (coming soon)

Local development

npm install
npm start      # Open http://localhost:3013
npm run build  # Static build in build/

Both start and serve are pinned to port 3013.

Refreshing the HTTP API reference

The /api-reference/http/* pages are generated from the OpenAPI spec vendored under vendor/atomicmemory-core-openapi.yaml. After atomicmemory-core ships a new release:

npm install @atomicmemory/atomicmemory-core@<new-version>
npm run vendor:spec     # refresh the vendored openapi.yaml
npm run regen:api       # clean + regenerate the .mdx artifacts
git commit -am "Refresh vendored core openapi spec"

prestart and prebuild both run regen:api so local dev and CI builds always render from the committed vendored spec.

Contributing

platform/ and sdk/ pages are authored by hand. api-reference/http/ pages are generated — edit atomicmemory-core/src/schemas/*.ts and regenerate via the workflow above to change them.

MIT-licensed. See atomicmemory-core for the engine this documents.