{
  "id": "page:docs-supermemory-research-raw-09-github-readme",
  "_kind": "Page",
  "_file": "wiki/docs/supermemory-research/raw/09-github-readme.md",
  "_cluster": "wiki",
  "attributes": {
    "nodeKind": "Page",
    "sourcePath": "docs/supermemory-research/raw/09-github-readme.md",
    "sourceKind": "repo-docs",
    "title": "Supermemory GitHub Repository",
    "displayName": "Supermemory GitHub Repository",
    "slug": "docs/supermemory-research/raw/09-github-readme",
    "articlePath": "wiki/docs/supermemory-research/raw/09-github-readme.md",
    "article": "\n# Supermemory GitHub Repository\n\nSource: https://github.com/supermemoryai/supermemory\n\n## Overview\n\nSupermemory is a \"memory engine and app that is extremely fast, scalable\" -- \"the Memory API for the AI era.\" Ranks #1 across three major benchmarks: LongMemEval, LoCoMo, and ConvoMem.\n\n## Core Features\n\n### Memory System\n\nAutomatically extracts facts from conversations, handles temporal changes and contradictions, implements automatic forgetting of expired information.\n\n### User Profiles\n\nDual-layer context: stable facts + recent activity. Retrievable in ~50ms per query.\n\n### Hybrid Search\n\nCombines RAG (document retrieval) with personalized memory in single queries.\n\n### Connectors\n\nIntegrates with Google Drive, Gmail, Notion, OneDrive, GitHub, and web crawlers with real-time webhook synchronization.\n\n### Multi-modal Processing\n\nPDFs, images (OCR), videos (transcription), code (AST-aware parsing).\n\n## Architecture\n\nFour layers:\n1. Memory Engine (fact extraction and contradiction resolution)\n2. User Profiles (static + dynamic context)\n3. Hybrid Search (combined RAG and memory)\n4. Connectors and File Processing\n\n## Installation\n\n```bash\nnpm install supermemory  # JavaScript/Node.js\npip install supermemory  # Python\n```\n\n## Integration Frameworks\n\nVercel AI SDK, LangChain, LangGraph, OpenAI Agents SDK, Mastra, Agno, Claude Memory Tool, n8n.\n\n## Client Plugins\n\n- Claude Supermemory Plugin: https://github.com/supermemoryai/claude-supermemory\n- OpenClaw Plugin: https://github.com/supermemoryai/openclaw-supermemory\n- OpenCode Plugin: https://github.com/supermemoryai/opencode-supermemory\n- Hermes Agent: https://github.com/NousResearch/hermes-agent\n\n## MemoryBench\n\nOpen-source benchmarking framework supporting Supermemory, Mem0, and Zep providers. Includes MemScore metric for comparing quality, latency, and token efficiency.\n\n## Tech Stack\n\nTypeScript (64%), MDX (28.9%), Python (6.2%). Built with Postgres, Remix, TailwindCSS, Vite, Cloudflare Workers/KV/Pages, Drizzle ORM.\n\n## License\n\nMIT\n",
    "documents": []
  },
  "outgoingEdges": [],
  "incomingEdges": [
    {
      "from": "page:docs-supermemory-research",
      "to": "page:docs-supermemory-research-raw-09-github-readme",
      "kind": "contains_page"
    }
  ]
}