# AtomicMemory

## Docs

- [AtomicMemory for LangChain JS: Memory Tools and Helpers](https://enterprise.atomicmemory.ai/adapters/langchain.md): Add durable semantic memory to any LangChain JS agent or chain using createMemoryTools, searchMemory, and ingestTurn from the AtomicMemory LangChain adapter.
- [AtomicMemory for LangGraph JS: Node Factories](https://enterprise.atomicmemory.ai/adapters/langgraph.md): Add durable memory to LangGraph JS graphs using createMemoryRetrieveNode and createMemoryIngestNode — typed graph nodes that slot into any StateGraph.
- [AtomicMemory for Mastra: Agent Tools and Helpers](https://enterprise.atomicmemory.ai/adapters/mastra.md): Add durable semantic memory to any Mastra agent using createMemoryTools, searchMemory, and ingestTurn from the AtomicMemory Mastra adapter.
- [AtomicMemory for OpenAI Agents SDK: Memory Primitives](https://enterprise.atomicmemory.ai/adapters/openai-agents.md): Wire durable semantic memory into OpenAI Agents SDK runs using runWithMemory, augmentInputWithMemory, ingestAgentTurn, and createMemoryTools.
- [AtomicMemory for Vercel AI SDK: Composable Primitives](https://enterprise.atomicmemory.ai/adapters/vercel-ai.md): Give any Vercel AI SDK message-driven model call durable semantic memory using withMemory, retrieve, augmentWithMemory, ingestTurn, and fromModelMessages.
- [API Authentication: Bearer Tokens and Key Management](https://enterprise.atomicmemory.ai/api/authentication.md): All AtomicMemory Core API requests require a Bearer token in the Authorization header. Learn how keys work and how to keep them secure.
- [POST /v1/memories/ingest — Store Memories with Extraction](https://enterprise.atomicmemory.ai/api/ingest.md): Two ingest endpoints let you store memories from conversations: full LLM extraction with AUDN-SC mutation, or fast embedding-based deduplication only.
- [Memory Management: List, Get, and Delete Operations](https://enterprise.atomicmemory.ai/api/memories.md): Use the list, get, and delete endpoints to browse memories by scope, fetch a single record by ID, or soft-delete memories from the active store.
- [AtomicMemory REST API: Endpoints, Auth, and Base URLs](https://enterprise.atomicmemory.ai/api/overview.md): The AtomicMemory Core REST API exposes ingest, search, memory management, and consolidation endpoints over HTTP/JSON from a single base URL.
- [POST /v1/memories/consolidate — Compress Context](https://enterprise.atomicmemory.ai/api/package.md): The consolidate endpoint compresses memories in a scope into a token-budgeted context string, ready to inject as a system message in your model call.
- [POST /v1/memories/search — Semantic Memory Retrieval](https://enterprise.atomicmemory.ai/api/search.md): Two search endpoints give you hybrid vector-plus-keyword retrieval or fast vector-only search, both scoped to a user, agent, namespace, or thread.
- [AtomicMemory CLI Command Reference](https://enterprise.atomicmemory.ai/cli/commands.md): Complete reference for every AtomicMemory CLI command, organized by group, with flags, usage notes, and copy-ready examples.
- [AtomicMemory CLI: Command-Line Memory Management](https://enterprise.atomicmemory.ai/cli/overview.md): Install and use the AtomicMemory CLI to manage memories, diagnose your setup, and integrate memory workflows into automation scripts and agent pipelines.
- [AtomicMemory Core Concepts: Scope, Ingest, and Retrieval](https://enterprise.atomicmemory.ai/core-concepts.md): Learn the key ideas behind AtomicMemory: how scope works, how memories are ingested and mutated, and how retrieval returns ranked results.
- [AtomicMemory FAQ: Architecture, Providers, and Integrations](https://enterprise.atomicmemory.ai/faq.md): Answers to the most common questions about AtomicMemory — architecture, providers, integrations, storage, and how the memory engine works.
- [Memory Audit Log: Tamper-Evident Operation History](https://enterprise.atomicmemory.ai/governance/audit.md): Track every operator action on agent memories through a tamper-evident receipt chain — who did what, when, and to which memory record.
- [Inspect and Correct Agent Memories in AtomicMemory](https://enterprise.atomicmemory.ai/governance/inspect-correct.md): How to find, review, and correct individual memories through the governance UI and the operator action API using your CORE_ADMIN_API_KEY.
- [Memory Governance: Inspect, Correct, Audit, and Control](https://enterprise.atomicmemory.ai/governance/overview.md): A tour of AtomicMemory's governance surfaces — inspect memories, correct wrong claims, resolve conflicts, and audit every operator action.
- [AtomicMemory: Semantic Memory for AI Agents and Apps](https://enterprise.atomicmemory.ai/introduction.md): AtomicMemory is a portable, inspectable semantic memory layer that captures context, grounds AI generations, and carries knowledge across sessions.
- [AtomicMemory MCP Server: Memory Tools for Any Agent](https://enterprise.atomicmemory.ai/mcp/overview.md): Run the AtomicMemory MCP server to expose memory_search, memory_ingest, memory_package, and memory_list as MCP tools to any compatible agent host.
- [AtomicMemory for Claude Code: Persistent Session Memory](https://enterprise.atomicmemory.ai/plugins/claude-code.md): Add persistent semantic memory to Claude Code: install the AtomicMemory plugin for an MCP server, automatic lifecycle hooks, and a built-in agent skill.
- [AtomicMemory for OpenAI Codex: Marketplace Memory Plugin](https://enterprise.atomicmemory.ai/plugins/codex.md): Install the AtomicMemory plugin for OpenAI Codex via the plugin marketplace to add persistent semantic memory, MCP tools, and optional lifecycle hooks.
- [AtomicMemory for Cursor: MCP Server Memory Integration](https://enterprise.atomicmemory.ai/plugins/cursor.md): Connect AtomicMemory to Cursor via a manual MCP config and project rule, giving the Cursor agent persistent semantic memory across your projects.
- [AtomicMemory for Hermes Agent: Native Memory Provider](https://enterprise.atomicmemory.ai/plugins/hermes.md): Install AtomicMemory as a native Hermes memory provider with background recall, turn sync, a circuit breaker, and configurable memory scope and mode.
- [AtomicMemory for Langflow: Custom Memory Components](https://enterprise.atomicmemory.ai/plugins/langflow.md): Install AtomicMemory's Langflow custom components to add persistent memory, governance inspection, and audit tools to any Langflow flow.
- [Get Started with AtomicMemory in 5 Minutes](https://enterprise.atomicmemory.ai/quickstart.md): Run AtomicMemory Core in Docker, install the TypeScript SDK, then ingest a conversation and search it by semantic query — all in under five minutes.
- [MemoryClient: Initialize and Use AtomicMemory in Code](https://enterprise.atomicmemory.ai/sdk/memory-client.md): Full reference for MemoryClient — initialization, ingest, semantic search, get, delete, package, and list with TypeScript examples.
- [AtomicMemory TypeScript SDK: Overview and Installation](https://enterprise.atomicmemory.ai/sdk/overview.md): Install @atomicmemory/sdk for a backend-agnostic memory layer with pluggable providers, local embeddings, storage adapters, and semantic search primitives.
- [SDK Providers: AtomicMemory Core and Mem0 Backends](https://enterprise.atomicmemory.ai/sdk/providers.md): Configure AtomicMemory Core or Mem0 as the SDK memory backend — swap providers with one config change and no rewrites to your application code.
- [Memory Scope: Routing and Isolating Agent Memories](https://enterprise.atomicmemory.ai/sdk/scope.md): Understand how scope fields — user, agent, namespace, and thread — route and isolate memories across users, agents, apps, and conversation sessions.
- [Configure AtomicMemory Core: Keys, Ports, and Storage](https://enterprise.atomicmemory.ai/self-hosting/configuration.md): Reference for AtomicMemory Core environment variables — API keys, embedding and LLM providers, port, database URL, and storage settings for production.
- [Self-Host AtomicMemory Core with Docker](https://enterprise.atomicmemory.ai/self-hosting/docker.md): Run AtomicMemory Core with Docker in minutes — persistent embedded Postgres, pluggable embedding and LLM providers, and a single docker run command.
- [Choose Embedding and LLM Providers for AtomicMemory](https://enterprise.atomicmemory.ai/self-hosting/providers.md): Pick embedding and LLM backends for AtomicMemory Core — local WASM, OpenAI, Anthropic, Ollama, Groq, Voyage AI, and any OpenAI-compatible endpoint.
- [Troubleshooting AtomicMemory: Common Issues and Fixes](https://enterprise.atomicmemory.ai/troubleshooting.md): Diagnose and fix the most common problems you may encounter when setting up and running AtomicMemory — from Docker ports to SDK initialization.