<!--
Full-page Markdown export (rendered HTML → GFM).
Source: https://neotoma.io/docs
Generated: 2026-05-04T09:49:43.295Z
-->
# Documentation
All Neotoma documentation organized by category. Start with getting started if you are new, or jump to the reference section for API, MCP, and CLI details.
## Getting started
- [
EvaluateHave your agent read this page to decide whether Neotoma fits your workflow
](/evaluate)
- [
InstallInstall and initialize Neotoma locally
](/install)
- [
What to store firstPick the first durable facts, commitments, and source-backed records to persist
](/what-to-store)
- [
Backup and restoreProtect the SQLite database, source files, and reconstruction history
](/backup)
- [
Expose tunnelUse HTTPS tunnels when remote MCP clients cannot launch local stdio
](/tunnel)
- [
WalkthroughEnd-to-end example across operating, building, and debugging
](/walkthrough)
## Integrations
- [
ChatGPTDeterministic memory for ChatGPT conversations
](/neotoma-with-chatgpt)
- [
ClaudeStructured state alongside Claude platform memory
](/neotoma-with-claude)
- [
Claude CodePersistent memory for Claude Code's local CLI agent
](/neotoma-with-claude-code)
- [
CodexCross-task memory and CLI fallback
](/neotoma-with-codex)
- [
CursorPersistent memory alongside Cursor context
](/neotoma-with-cursor)
- [
IronClawStructured MCP memory for IronClaw agents
](/neotoma-with-ironclaw)
- [
OpenClawUser-owned memory for OpenClaw agents
](/neotoma-with-openclaw)
- [
OpenCodeLifecycle hooks and MCP memory for OpenCode
](/neotoma-with-opencode)
## Reference
- [
InstallInstall and initialize Neotoma locally
](/install)
- [
REST APIOpenAPI endpoints and parameters
](/api)
- [
MCP serverModel Context Protocol actions
](/mcp)
- [
CLICommands, flags, and REPL
](/cli)
- [
Memory guaranteesAll memory properties on one page
](/memory-guarantees)
- [
Memory modelsPlatform, retrieval, file-based, and deterministic memory compared
](/memory-models)
- [
FoundationsPrivacy-first architecture and cross-platform design
](/foundations)
- [
Agent instructionsMandatory behavioral rules for agents using Neotoma
](/agent-instructions)
- [
ArchitectureState flow, guarantees, and principles
](/architecture)
- [
TerminologyGlossary of key concepts
](/terminology)
- [
TroubleshootingCommon failure modes and practical fixes
](/troubleshooting)
- [
ChangelogRelease history and documentation updates
](/changelog)
- [
All pages (Markdown)Every indexable route as Markdown (SEO summaries)
](/site-markdown)
## Primitive record types
- [
OverviewThe seven system-level building blocks behind every entity, snapshot, and audit trail
](/primitives)
- [
EntitiesCanonical row for every person, company, or thing, deterministic ID, aliases, and merge tracking
](/primitives/entities)
- [
Entity snapshotsReducer output with per-field provenance back to observations and an optional embedding column
](/primitives/entity-snapshots)
- [
SourcesContent-addressed raw storage with SHA-256 deduplication per user
](/primitives/sources)
- [
InterpretationsVersioned, audited extraction attempts with full interpretation\_config provenance
](/primitives/interpretations)
- [
ObservationsGranular, immutable facts that the reducer composes into entity snapshots
](/primitives/observations)
- [
RelationshipsFirst-class typed graph edges that follow the same observation-snapshot pattern
](/primitives/relationships)
- [
Timeline eventsSource-anchored temporal records derived deterministically from extracted dates
](/primitives/timeline-events)
## Schemas
- [
OverviewVersioned, config-driven definitions that give the immutable primitives their domain shape
](/schemas)
- [
Schema registryThe table that holds every versioned schema\_definition + reducer\_config, global or per-user
](/schemas/registry)
- [
Merge policiesPer-field declarative rules, last\_write, highest\_priority, most\_specific, merge\_array
](/schemas/merge-policies)
- [
Storage layersraw\_text, properties, and raw\_fragments, the three places extracted data can land
](/schemas/storage-layers)
- [
Versioning & evolutionSemver, additive minor bumps, breaking major bumps, and the public schema snapshots dump
](/schemas/versioning)
- [
Schema management (CLI)CLI workflows for listing, validating, evolving, and registering schemas at runtime
](/schema-management)
## Where the tax shows up
- [
Context janitorYou re-explain context every session. State that persists across tools.
](/operating)
- [
Inference varianceCorrections don’t stick. Deterministic state for agent pipelines.
](/building-pipelines)
- [
Log archaeologySame inputs, different state. Replayable timelines and diffs.
](/debugging-infrastructure)
## Compare
- [
Build vs buyWhen to adopt a state-integrity layer instead of building around observability alone
](/build-vs-buy)
- [
Neotoma vs platform memoryConvenience inside one AI product versus portable, auditable state across tools
](/neotoma-vs-platform-memory)
- [
Neotoma vs Mem0Retrieval memory for prompt augmentation versus deterministic entity state
](/neotoma-vs-mem0)
- [
Neotoma vs ZepKnowledge-graph retrieval versus versioned, schema-bound state
](/neotoma-vs-zep)
- [
Neotoma vs RAGRelevant chunk retrieval versus exact state reconstruction
](/neotoma-vs-rag)
- [
Neotoma vs file-based memoryMarkdown and JSON portability versus structured guarantees and provenance
](/neotoma-vs-files)
- [
Neotoma vs database memoryCRUD rows versus append-only observations and deterministic reducers
](/neotoma-vs-database)
## External
- [
GitHub repositorySource code, README, and issues
](https://github.com/markmhendrickson/neotoma)
- [
npm packageInstall via npm
](https://www.npmjs.com/package/neotoma)