{
"markdown": "# AutoMem MCP: Give Your AI Perfect Memory\n\n
\n \n
\n\n
\n \n \n \n \n
\n\n**One command. Infinite memory. Perfect recall across all your AI tools.**\n\n```bash\nnpx @verygoodplugins/mcp-automem setup\n```\n\nYour AI assistant now remembers everything. Forever. Across every conversation.\n\n
The guided installer โ npx @verygoodplugins/mcp-automem install walks you through local, hosted, or existing-endpoint setup.
\n\nWorks with **Claude Desktop**, **Cursor IDE**, **Claude Code**, **GitHub Copilot (coding agent)**, **ChatGPT**, **ElevenLabs**, **OpenAI Codex**, **OpenClaw**, **Hermes**, **Grok Build**, **Google Antigravity** - any MCP-compatible AI platform.\n\n## The Problem We Solve\n\nEvery AI conversation starts from zero. Claude forgets your coding style. Cursor can't learn your patterns. Your assistant doesn't remember yesterday's decisions.\n\n**Until now.**\n\nAutoMem MCP connects your AI to persistent memory powered by **[AutoMem](https://github.com/verygoodplugins/automem)** - a graph-vector memory service.\n\n## What You Get\n\n### ๐ง Persistent Memory Across Sessions\n\n- AI remembers decisions, patterns, and context **forever**\n- Works across **all MCP platforms** - Claude Desktop, Cursor, Claude Code, OpenAI Codex, OpenClaw, Hermes, Grok Build, Google Antigravity\n- **Cross-device sync** - same memory on Mac, Windows, Linux\n\n### ๐ Graph-Vector Architecture\n\n- **11 public authorable relationship types** between memories (recall results may also include read-only system/internal relations that are not valid `associate_memories` inputs)\n- **Research-validated** approach (HippoRAG 2: 7% better associative memory)\n- **Sub-second retrieval** even with millions of memories\n\n### ๐ Works Everywhere You Code\n\n| Platform | Support | Setup Time |\n| ------------------ | ------- | ---------- |\n| **Claude Desktop** | โ Full | 30 seconds |\n| **Cursor IDE** | โ Full | 30 seconds |\n| **Claude Code** | โ Full | 30 seconds |\n| **GitHub Copilot** | โ Full | 2 minutes |\n| **OpenAI Codex** | โ Full | 30 seconds |\n| **OpenClaw** | โ Full | 30 seconds |\n| **Hermes Agent** | โ Full | 30 seconds |\n| **Grok Build** | โ Full | 30 seconds |\n| **Google Antigravity** | โ Full | 30 seconds |\n| **Any MCP client** | โ Full | 30 seconds |\n\n## See It In Action\n\n### Claude Desktop with Personal Preferences\n\n\n_Claude automatically recalls memories using the Personal Preferences template_\n\n### Cursor IDE with Memory Rules\n\n\n_Cursor uses automem.mdc rule to automatically recall and store memories_\n\n### Claude Code with Session Memory\n\n\n_Session-start recall plus LLM-judged storage: Claude decides what's durable and stores it via the memory tools_\n\nMore platform walkthroughs (Codex, Hermes, Antigravity, remote MCP) live in the **[Installation Guide](INSTALLATION.md)**.\n\n## Quick Start\n\n### 1. Set Up AutoMem Service\n\nYou need a running AutoMem service (the memory backend). Choose one:\n\n**Option A: Local Development** (fastest, free)\n\n```bash\ngit clone https://github.com/verygoodplugins/automem.git\ncd automem\nmake dev\n```\n\nService runs at `http://localhost:8001` - perfect for single-machine use.\n\n**Option B: Railway Cloud** (recommended for production)\n\n[](https://github.com/verygoodplugins/automem/blob/main/INSTALLATION.md#railway-deployment)\n\nOne-click deploy with $5 free credits. Typical cost: ~$0.50-1/month after trial.\n\n๐ **[AutoMem Service Installation Guide](https://github.com/verygoodplugins/automem/blob/main/INSTALLATION.md)** - Complete setup instructions for local, Railway, Docker, and production deployments.\n\n---\n\n### 2. Install MCP Client\n\n#### Claude Desktop - One-Click Install\n\nDownload and double-click to install AutoMem in Claude Desktop:\n\n**[โฌ๏ธ Download AutoMem for Claude Desktop (.mcpb)](https://github.com/verygoodplugins/mcp-automem/releases/latest/download/mcp-automem.mcpb)**\n\nAfter installing:\n1. Claude Desktop will prompt you for your **AutoMem Endpoint** (`http://127.0.0.1:8001` for local)\n2. Optionally enter your **API Key** (required for Railway, skip for local)\n3. Click Enable\n\nThen add the paste-ready Personal Preferences starter from [`templates/CLAUDE_DESKTOP_INSTRUCTIONS.md`](templates/CLAUDE_DESKTOP_INSTRUCTIONS.md). That's it: Claude now has persistent memory and knows when to use it.\n\n#### Other Platforms\n\nConnect your AI tools to the AutoMem service you just started.\n\n```bash\n# Guided install - pick where AutoMem runs, verify it, write .env, and\n# configure your agents (Codex, Claude Code, Cursor, OpenClaw, Hermes)\nnpx @verygoodplugins/mcp-automem install\n```\n\nEvery change is shown in a review plan before anything is written, and each\nmodified file keeps a `.bak` backup. Add `--dry-run` to preview, `--yes` to\napply non-interactively. See the [Installation Guide](INSTALLATION.md#guided-install-fastest)\nfor all flags.\n\nJust need the `.env` + config snippets without the agent setup? Use the lighter wizard:\n\n```bash\n# Creates .env and prints config for your AI platform\nnpx @verygoodplugins/mcp-automem setup\n```\n\n**When prompted:**\n\n- **AutoMem Endpoint:** `http://localhost:8001` (or your Railway URL if deployed)\n- **API Key:** Leave blank for local development (or paste your token for Railway)\n\nThe wizard will:\n\n- โ Save your endpoint and API key to `.env`\n- โ Generate config snippets for Claude Desktop/Cursor/Code\n- โ Validate connection to your AutoMem service\n\n### 3. Platform-Specific Setup\n\n**For Claude Code (plugin โ recommended):**\n\n```text\n# In Claude Code:\n/plugin marketplace add verygoodplugins/mcp-automem\n/plugin install automem@verygoodplugins-mcp-automem\n```\n\nClaude Code prompts for your AutoMem URL and API key at enable time, bundles the MCP server and silent recall/store-tracking hooks, and auto-updates. Prefer hooks and permissions written directly into `~/.claude/` instead? Run `npx @verygoodplugins/mcp-automem claude-code`.\n\nOn Windows, the hook payload assumes a POSIX shell environment such as Git Bash, MSYS2, or WSL โ only `bash` is required (the hooks are pure bash+sed).\n\n**For Cursor IDE:**\n\n[](cursor://anysphere.cursor-deeplink/mcp/install?name=memory&config=eyJlbnYiOnsiQVVUT01FTV9BUElfVVJMIjoiaHR0cDovLzEyNy4wLjAuMTo4MDAxIiwiQVVUT01FTV9BUElfS0VZIjoieW91ci1hcGkta2V5LWlmLXJlcXVpcmVkIn0sImNvbW1hbmQiOiJucHggLXkgQHZlcnlnb29kcGx1Z2lucy9tY3AtYXV0b21lbSJ9)\n\n```bash\n# Or use CLI to install automem.mdc rule file\nnpx @verygoodplugins/mcp-automem cursor\n```\n\n**Other platforms** โ Claude Desktop (one-click `.mcpb` above, plus the [Personal Preferences template](templates/CLAUDE_DESKTOP_INSTRUCTIONS.md)), [OpenAI Codex](INSTALLATION.md#openai-codex), [Hermes Agent](INSTALLATION.md#hermes-agent), [Grok Build](INSTALLATION.md#grok-build), [OpenClaw](INSTALLATION.md#openclaw), [Google Antigravity](INSTALLATION.md#google-antigravity), and [GitHub Copilot](INSTALLATION.md#github-copilot-coding-agent-githubcom):\n\n๐ **[Full Installation Guide](INSTALLATION.md)** for every platform's setup and verification steps\n\n---\n\n## Remote MCP via HTTP\n\nAn optional sidecar service (deployable to Railway or any Docker host) connects AutoMem to platforms that support remote MCP over **Streamable HTTP** or SSE โ ChatGPT (Developer Mode connectors), Claude.ai web and Claude Mobile, and ElevenLabs Agents.\n\n๐ **[Remote MCP setup](INSTALLATION.md#remote-mcp-via-http-sidecar)** for deployment, connect URLs, and per-platform screenshots.\n\n## Architecture\n\n```\nโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\nโ Your AI Platforms โ\nโ Claude Desktop โ Cursor โ Claude Code โ\nโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\n โ MCP Protocol\n โผ\nโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\nโ @verygoodplugins/mcp-automem (this repo) โ\nโ โข Translates MCP calls โ AutoMem API โ\nโ โข Platform integrations & rules โ\nโ โข Handles authentication โ\nโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\n โ HTTP API\n โผ\nโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\nโ AutoMem Service (separate repo) โ\nโ github.com/verygoodplugins/automem โ\nโ โโโโโโโโโโโโโโ โโโโโโโโโโโโโโ โ\nโ โ FalkorDB โ โ Qdrant โ โ\nโ โ (Graph) โ โ (Vectors) โ โ\nโ โโโโโโโโโโโโโโ โโโโโโโโโโโโโโ โ\nโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ\n```\n\n**This repo (mcp-automem):**\n\n- MCP client that connects AI platforms to AutoMem\n- Platform-specific integrations (Cursor rules, Claude Code hooks, etc.)\n- Setup wizards and configuration tools\n\n**[AutoMem service](https://github.com/verygoodplugins/automem):**\n\n- Backend memory service with graph + vector storage\n- Deployment guides (local, Railway, Docker, production)\n- API server with FalkorDB + Qdrant\n\n## Features\n\n### Core Memory Operations\n\n- **`store_memory`** โ Save memories with content, tags, importance, metadata. Two modes:\n - **Single (default)**: top-level `content` plus optional fields, including `embedding`, `t_valid`, `t_invalid`, custom `id`.\n - **Batch**: `memories: [...]` (โค500 items) for bulk ingestion. Per-item `id`/`embedding`/`t_valid`/`t_invalid` are not supported in batch mode.\n- **`recall_memory`** โ Three modes selected by which params you pass:\n - **ID fetch**: `memory_id` โ fetches one memory by ID; updates `last_accessed`.\n - **Tag enumeration**: `tags` + `exhaustive: true` โ paginated exact-match listing for cleanup/audit workflows where ranked recall undercounts. Pair with `limit` (โค200) and `offset`; returns `has_more`.\n - **Ranked retrieval (default)**: hybrid search across vector, keyword, tags, recency/state controls, score filters, and graph expansion. Supports `state_mode`, `recency_bias`, `scope_fallback`, `expand_respect_tags`, `min_score`, `adaptive_floor`, and diagnostics such as `tag_scope`, `score_filter`, `query_time_ms`, `vector_search`, and per-result `outside_tag_scope`/`state_replaces`.\n- **`associate_memories`** โ Create relationships (11 public authorable types; recall results may also include read-only system relations). Supports single-pair mode and batch mode via `associations: [...]` (โค500) with relation-specific props like `reason`, `context`, `resolution`, `observations`, `transformation`, and `role`.\n- **`update_memory`** โ Modify existing memories\n- **`delete_memory`** โ Two modes:\n - **Single (default)**: `memory_id` โ removes one memory and its embedding.\n - **Bulk-by-tag**: `tags: [...]` โ bulk-delete all memories matching ANY tag (exact, case-insensitive). No dry-run; verify with `recall_memory({ tags, exhaustive: true })` first.\n- **`check_database_health`** โ Monitor service health, degraded state, sync counts, vector dimensions, and enrichment diagnostics when the service provides them\n\n### Advanced Recall (v0.8.0+)\n\n**Multi-hop Reasoning** - Answer complex questions like \"What is Amanda's sister's career?\"\n\n```javascript\nmcp__memory__recall_memory({\n query: \"What is Amanda's sister's career?\",\n expand_entities: true, // Finds \"Amanda's sister is Rachel\" โ memories about Rachel\n});\n```\n\n**Context-Aware Coding** - Recall prioritizes language and style preferences\n\n```javascript\nmcp__memory__recall_memory({\n query: \"error handling patterns\",\n language: \"typescript\",\n context_types: [\"Style\", \"Pattern\"],\n});\n```\n\n### Platform Integrations\n\n#### Cursor IDE\n\n- โ **Memory-first rule file** (`automem.mdc` in `.cursor/rules/`)\n- โ **Automatic memory recall** at conversation start\n- โ **Auto-detects project context** (package.json, git remote)\n- โ **Global user rules option** for all projects\n- โ **Simple setup** via CLI or one-click install\n\n#### Claude Code\n\n- โ **Native plugin** - MCP server, silent hooks, and skill in one `/plugin install`, with enable-time config prompts and auto-updates\n- โ **LLM-judged storage** - session-start guidance nudges Claude to store, verify, and associate durable memories during normal work\n- โ **Memory rules** in CLAUDE.md guide Claude's memory usage\n\n#### GitHub Copilot\n\n- โ **Standalone hook JSON files** installed into `$COPILOT_HOME/hooks/` or `~/.copilot/hooks/`\n- โ **Memory rules** template for `copilot-instructions.md`\n- โ **Format flag** - `--format cli` (camelCase) or `--format vscode` (PascalCase)\n- โ **Setup**: `npx @verygoodplugins/mcp-automem copilot --yes`\n\n#### Claude Desktop\n\n- โ Direct MCP integration\n- โ Paste-ready Personal Preferences starter template\n- โ Full memory API access\n\n## Why AutoMem MCP?\n\n### vs. Building Your Own\n\n- โ **2 years of R&D** already done\n- โ **Research-validated** architecture (HippoRAG 2, MELODI, A-MEM)\n- โ **Working integrations** across all MCP platforms\n- โ **Active development** and community\n\n### vs. Other Memory Solutions\n\n- โ **True graph relationships** (not just vector similarity)\n- โ **Universal MCP compatibility** (works with any MCP client)\n- โ **7 memory types** (Decision/Pattern/Preference/Style/Habit/Insight/Context)\n- โ **Self-hostable** ($5/month vs $150+ for alternatives)\n\n### vs. Native AI Memory\n\n- โ **Persistent across sessions** (not just context window)\n- โ **Cross-platform** (same memory in Claude, Cursor, Code)\n- โ **Structured relationships** (not just RAG)\n- โ **Infinite scale** (no context window limits)\n\n## Documentation\n\n### MCP Client & Integrations (this repo)\n\n- ๐ฆ **[Installation Guide](INSTALLATION.md)** - MCP client setup for all platforms\n- ๐ **[Remote MCP via HTTP](INSTALLATION.md#remote-mcp-via-http-sidecar)** - Connect ChatGPT, Claude Web/Mobile, ElevenLabs\n- ๐ฏ **[Cursor Setup](INSTALLATION.md#cursor-ide)** - IDE integration with rules\n- ๐ค **[Claude Code Setup](templates/CLAUDE_CODE_INTEGRATION.md)** - Plugin install, hooks, and memory rules\n- โ ๏ธ **[Deprecations](DEPRECATION.md)** - History of the plugin deprecation and its reversal\n- ๐ **[OpenAI Codex Setup](INSTALLATION.md#openai-codex)** - Codex CLI/IDE/Cloud integration\n- ๐ช **[Google Antigravity Setup](INSTALLATION.md#google-antigravity)** - Raw MCP config via Antigravity's MCP Store\n- ๐ **[MCP Tools Reference](INSTALLATION.md#mcp-tools)** - All memory operations\n- ๐ **[Changelog](CHANGELOG.md)** - Release history\n\n### AutoMem Service (separate repo)\n\n- ๐๏ธ **[AutoMem Service](https://github.com/verygoodplugins/automem)** - Backend repository\n- ๐ **[Service Installation](https://github.com/verygoodplugins/automem/blob/main/INSTALLATION.md)** - Local, Railway, Docker deployment\n- โ๏ธ **[API Documentation](https://github.com/verygoodplugins/automem#api-reference)** - REST API reference\n- ๐งช **[Evaluation Lab](https://github.com/verygoodplugins/automem-evals)** - Exploratory recall-quality benchmarks and ruleset A/B testing\n\n## The Science Behind AutoMem\n\nThe AutoMem service implements cutting-edge 2025 research:\n\n- **[HippoRAG 2](https://arxiv.org/abs/2502.14802)** (OSU, June 2025): Graph-vector approach achieves 7% better associative memory\n- **A-MEM** (July 2025): Dynamic memory organization with Zettelkasten principles\n- **MELODI** (DeepMind, 2025): 8x memory compression without quality loss\n- **ReadAgent** (DeepMind, 2024): 20x context extension through gist memories\n\nThis MCP package provides the bridge between your AI and that research-validated memory system. The backend has also been benchmarked on the neutral [Agent Memory Benchmark](https://automem.ai/benchmarks), including BEAM large-context scaling tiers โ reproducible end to end, so you can run it yourself.\n\n## Community & Support\n\n- ๐ฌ **[Discord](https://automem.ai/discord)** - Join the community, get help, share feedback\n- ๐ฆ **[X Community](https://x.com/i/communities/2013114118912225326)** - Discussion and updates\n- ๐ฃ **[@automem_ai](https://x.com/automem_ai)** - Official announcements\n- ๐ฆ **[NPM Package](https://www.npmjs.com/package/@verygoodplugins/mcp-automem)** - This MCP client\n- ๐ฌ **[AutoMem Service](https://github.com/verygoodplugins/automem)** - Backend repo with deployment guides\n- ๐ **[GitHub Issues](https://github.com/verygoodplugins/mcp-automem/issues)** - Bug reports and feature requests\n\n## Contributing\n\nWe welcome contributions! Please:\n\n1. Fork the repository\n2. Create a feature branch\n3. Make your changes with tests\n4. Submit a pull request with a Conventional Commit title such as `fix:`, `feat:`, `docs:`, or `chore:`\n5. Do not prefix the PR title with labels like `[codex]` or `[wip]` because the squash-merge commit is taken from the PR title\n\n## License\n\nMIT - Because great memory should be free.\n\n---\n\n**Ready to give your AI perfect memory?**\n\n```bash\nnpx @verygoodplugins/mcp-automem setup\n```\n\n_Built with obsession. Validated by neuroscience. Powered by graph theory. Works with every MCP-enabled AI._\n\n_Designed by Jack Arturo at [Very Good Plugins](https://verygoodplugins.com)_ ๐งก\n\n**Transform your AI from a tool into a teammate. Start now.**\n",
"bytes": 17937,
"sha": "0bf4ec4272eaa3efba16adb21611d6160746265d1dc043026da01b67a04b0d7e",
"repo_slug": "verygoodplugins/mcp-automem",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/mcp_io_github_verygoodplugins_mcp_automem_42ea96d7/readme"
}
![\"AutoMem\"](\"assets/icon.svg\"){kind=icon}
\n![\"Version\"](\"https://img.shields.io/npm/v/@verygoodplugins/mcp-automem\")
![\"License\"](\"https://img.shields.io/npm/l/@verygoodplugins/mcp-automem\")
![\"Discord\"](\"https://img.shields.io/badge/Discord-Join%20Community-5865F2?logo=discord&logoColor=white\"){kind=icon}
![\"X](\"https://img.shields.io/badge/X-@automem__ai-000000?logo=x&logoColor=white\"){kind=icon}