[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-2677":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":47,"readmeContent":48,"aiSummary":49,"trendingCount":16,"starSnapshotCount":16,"syncStatus":50,"lastSyncTime":51,"discoverSource":52},2677,"cognee","topoteretes\u002Fcognee","topoteretes","Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.","https:\u002F\u002Fwww.cognee.ai",null,"Python",17787,1883,73,18,0,28,118,607,117,119.83,"Apache License 2.0",false,"main",true,[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46],"ai","ai-agents","ai-memory","cognitive-architecture","cognitive-memory","context-engineering","contributions-welcome","good-first-issue","good-first-pr","graph-database","graph-rag","graphrag","help-wanted","knowledge","knowledge-graph","neo4j","open-source","openai","rag","vector-database","2026-06-12 04:00:15","\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\">\n    \u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Ftopoteretes\u002Fcognee\u002Frefs\u002Fheads\u002Fdev\u002Fassets\u002Fcognee-logo-transparent.png\" alt=\"Cognee Logo\" height=\"60\">\n  \u003C\u002Fa>\n\n  \u003Cbr \u002F>\n\n  Cognee - The Brain behind your Agents\n\n  \u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=8hmqS2Y5RVQ&t=13s\">Demo\u003C\u002Fa>\n  .\n  \u003Ca href=\"https:\u002F\u002Fdocs.cognee.ai\u002F\">Docs\u003C\u002Fa>\n  .\n  \u003Ca href=\"https:\u002F\u002Fcognee.ai\">Learn More\u003C\u002Fa>\n  ·\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FNQPKmU5CCg\">Join Discord\u003C\u002Fa>\n  ·\n  \u003Ca href=\"https:\u002F\u002Fwww.reddit.com\u002Fr\u002FAIMemory\u002F\">Join r\u002FAIMemory\u003C\u002Fa>\n  .\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee-community\">Community Plugins & Add-ons\u003C\u002Fa>\n  \u003C\u002Fp>\n\n\n  [![GitHub forks](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Ftopoteretes\u002Fcognee.svg?style=social&label=Fork&maxAge=2592000)](https:\u002F\u002FGitHub.com\u002Ftopoteretes\u002Fcognee\u002Fnetwork\u002F)\n  [![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftopoteretes\u002Fcognee.svg?style=social&label=Star&maxAge=2592000)](https:\u002F\u002FGitHub.com\u002Ftopoteretes\u002Fcognee\u002Fstargazers\u002F)\n  [![GitHub commits](https:\u002F\u002Fbadgen.net\u002Fgithub\u002Fcommits\u002Ftopoteretes\u002Fcognee)](https:\u002F\u002FGitHub.com\u002Ftopoteretes\u002Fcognee\u002Fcommit\u002F)\n  [![GitHub tag](https:\u002F\u002Fbadgen.net\u002Fgithub\u002Ftag\u002Ftopoteretes\u002Fcognee)](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\u002Ftags\u002F)\n  [![Downloads](https:\u002F\u002Fstatic.pepy.tech\u002Fbadge\u002Fcognee)](https:\u002F\u002Fpepy.tech\u002Fproject\u002Fcognee)\n  [![License](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Ftopoteretes\u002Fcognee?colorA=00C586&colorB=000000)](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\u002Fblob\u002Fmain\u002FLICENSE)\n  [![Contributors](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcontributors\u002Ftopoteretes\u002Fcognee?colorA=00C586&colorB=000000)](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\u002Fgraphs\u002Fcontributors)\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsponsors\u002Ftopoteretes\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSponsor-❤️-ff69b4.svg\" alt=\"Sponsor\">\u003C\u002Fa>\n\n\u003Cp>\n  \u003Ca href=\"https:\u002F\u002Ftrendshift.io\u002Frepositories\u002F13955\" target=\"_blank\" style=\"display:inline-block;\">\n    \u003Cimg src=\"https:\u002F\u002Ftrendshift.io\u002Fapi\u002Fbadge\u002Frepositories\u002F13955\" alt=\"topoteretes%2Fcognee | Trendshift\" width=\"250\" height=\"55\" \u002F>\n  \u003C\u002Fa>\n\u003C\u002Fp>\n\nCognee gives AI agents a shared, improving memory of your data, decisions, and workflows so they can recall, connect, and act with context.\n\n  \u003Cp align=\"center\">\n  🌐 Available Languages\n  :\n  \u003C!-- Keep these links. Translations will automatically update with the README. -->\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=de\">Deutsch\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=es\">Español\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=fr\">Français\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=ja\">日本語\u003C\u002Fa> |\n  \u003Ca href=\"README_ko.md\">한국어\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=pt\">Português\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=ru\">Русский\u003C\u002Fa> |\n  \u003Ca href=\"https:\u002F\u002Fwww.readme-i18n.com\u002Ftopoteretes\u002Fcognee?lang=zh\">中文\u003C\u002Fa>\n  \u003C\u002Fp>\n\n\n\u003Cdiv style=\"text-align: center\">\n  \u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002Ftopoteretes\u002Fcognee\u002Frefs\u002Fheads\u002Fmain\u002Fassets\u002Fcognee_benefits.png\" alt=\"Why cognee?\" width=\"80%\" \u002F>\n\u003C\u002Fdiv>\n\u003C\u002Fdiv>\n\n\n\n\n## About Cognee\n\nCognee is an open-source memory control plane for your Agents that lets you ingest data in any format or structure and continuously learns to provide the right context. It combines embeddings, graphs and cognitive science approaches to make your documents both searchable by meaning and connected by relationships as they change and evolve.\n\n\n\n:star: _Help us reach more developers and grow the cognee community. Star this repo!_\n\n:books: _Check our detailed [documentation](https:\u002F\u002Fdocs.cognee.ai\u002Fgetting-started\u002Finstallation#environment-configuration) for setup and configuration._\n\n:crab: _Available as a plugin for your OpenClaw — [cognee-openclaw](https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002F@cognee\u002Fcognee-openclaw)_\n\n✴️ _Available as a plugin for your Claude Code — [claude-code-plugin](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee-integrations\u002Ftree\u002Fmain\u002Fintegrations\u002Fclaude-code)_\n\n\n\n### Why use Cognee:\n\n- Easily Build Company Brain - unify data from various sources in one place and enable Agents with your domain knowledge\n- Knowledge infrastructure — unified ingestion, graph\u002Fvector search, runs locally, ontology grounding, multimodal\n- Persistent and Learning Agents - learn from feedback, context management, cross-agent knowledge sharing\n- Reliable and Trustworthy Agents - agentic user\u002Ftenant isolation, traceability, OTEL collector, audit traits\n\n### Product Features\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Fcognee_products.png\" alt=\"Cognee Products\" width=\"80%\" \u002F>\n\u003C\u002Fp>\n\n## Basic Usage & Feature Guide\n\nTo learn more, [check out this short, end-to-end Colab walkthrough](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F12Vi9zID-M3fpKpKiaqDBvkk98ElkRPWy?usp=sharing) of Cognee's core features.\n\n[![Open In Colab](https:\u002F\u002Fcolab.research.google.com\u002Fassets\u002Fcolab-badge.svg)](https:\u002F\u002Fcolab.research.google.com\u002Fdrive\u002F12Vi9zID-M3fpKpKiaqDBvkk98ElkRPWy?usp=sharing)\n\n## Quickstart\n\nLet’s try Cognee in just a few lines of code.\n\n### Prerequisites\n\n- Python 3.10 to 3.14\n\n### Step 1: Install Cognee\n\nYou can install Cognee with **pip**, **poetry**, **uv**, or your preferred Python package manager.\n\n```bash\nuv pip install cognee\n```\n\n### Step 2: Configure the LLM\n```python\nimport os\nos.environ[\"LLM_API_KEY\"] = \"YOUR OPENAI_API_KEY\"\n```\nAlternatively, create a `.env` file using our [template](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\u002Fblob\u002Fmain\u002F.env.template).\n\nTo integrate other LLM providers, see our [LLM Provider Documentation](https:\u002F\u002Fdocs.cognee.ai\u002Fsetup-configuration\u002Fllm-providers).\n\n### Step 3: Run the Pipeline\n\nCognee's API gives you four operations — `remember`, `recall`, `forget`, and `improve`:\n\n```python\nimport cognee\nimport asyncio\n\n\nasync def main():\n    # Store permanently in the knowledge graph (runs add + cognify + improve)\n    await cognee.remember(\"Cognee turns documents into AI memory.\")\n\n    # Store in session memory (fast cache, syncs to graph in background)\n    await cognee.remember(\"User prefers detailed explanations.\", session_id=\"chat_1\")\n\n    # Query with auto-routing (picks best search strategy automatically)\n    results = await cognee.recall(\"What does Cognee do?\")\n    for result in results:\n        print(result)\n\n    # Query session memory first, fall through to graph if needed\n    results = await cognee.recall(\"What does the user prefer?\", session_id=\"chat_1\")\n    for result in results:\n        print(result)\n\n    # Delete when done\n    await cognee.forget(dataset=\"main_dataset\")\n\n\nif __name__ == '__main__':\n    asyncio.run(main())\n\n```\n\n### Use the Cognee CLI\n\n```bash\ncognee-cli remember \"Cognee turns documents into AI memory.\"\n\ncognee-cli recall \"What does Cognee do?\"\n\ncognee-cli forget --all\n```\n\nTo open the local UI, run:\n```bash\ncognee-cli -ui\n```\n\n## Use with AI Agents\n\n### Claude Code\n\nInstall the [Cognee memory plugin](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee-integrations\u002Ftree\u002Fmain\u002Fintegrations\u002Fclaude-code) to give Claude Code persistent memory across sessions. The plugin automatically captures tool calls into session memory via hooks and syncs to the permanent knowledge graph at session end.\n\n**Setup:**\n\n```bash\n# Install cognee\npip install cognee\n\n# Configure\nexport LLM_API_KEY=\"your-openai-key\"\n\n# Clone the plugin\ngit clone https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee-integrations.git\n\n# Enable it (add to ~\u002F.zshrc for permanent use)\nclaude --plugin-dir .\u002Fcognee-integrations\u002Fintegrations\u002Fclaude-code\n```\n\nOr connect to Cognee Cloud instead of running locally:\n\n```bash\nexport COGNEE_SERVICE_URL=\"https:\u002F\u002Fyour-instance.cognee.ai\"\nexport COGNEE_API_KEY=\"ck_...\"\n```\n\nThe plugin hooks into Claude Code's lifecycle — `SessionStart` initializes memory, `PostToolUse` captures actions, `UserPromptSubmit` injects relevant context, `PreCompact` preserves memory across context resets, and `SessionEnd` bridges session data into the permanent graph.\n\n### Hermes Agent\n\nEnable Cognee as the memory provider in [Hermes Agent](https:\u002F\u002Fgithub.com\u002FNousResearch\u002Fhermes-agent) for session-aware knowledge graph memory with auto-routing recall.\n\n**Setup:**\n\n```yaml\n# ~\u002F.hermes\u002Fconfig.yaml\nmemory:\n  provider: cognee\n```\n\n```bash\nexport LLM_API_KEY=\"your-openai-key\"\nhermes  # start chatting — session memory and graph persistence are automatic\n```\n\nOr run `hermes memory setup` and select Cognee. For Cognee Cloud, set `COGNEE_SERVICE_URL` and `COGNEE_API_KEY` in `~\u002F.hermes\u002F.env`.\n\n\n### Connect to Cognee Cloud\n\nPoint any Python agent at a managed Cognee instance — all SDK calls route to the cloud:\n\n```python\nimport cognee\n\nawait cognee.serve(url=\"https:\u002F\u002Fyour-instance.cognee.ai\", api_key=\"ck_...\")\n\nawait cognee.remember(\"important context\")\nresults = await cognee.recall(\"what happened?\")\n\nawait cognee.disconnect()\n```\n\n## Examples\n\nBrowse more examples in the [`examples\u002F`](examples\u002F) folder — demos, guides, custom pipelines, and database configurations.\n\n**Use Case 1 — Customer Support Agent**\n\n```python\nGoal: Resolve customer issues using their personal data across finance, support, and product history.\n\nUser: \"My invoice looks wrong and the issue is still not resolved.\"\n\nCognee tracks: past interactions, failed actions, resolved cases, product history\n\n# Agent response:\nAgent: \"I found 2 similar billing cases resolved last month.\n        The issue was caused by a sync delay between payment\n        and invoice systems — a fix was applied on your account.\"\n\n# What happens under the hood:\n- Unifies data sources from various company channels\n- Reconstructs the interaction timeline and tracks outcomes\n- Retrieves similar resolved cases\n- Maps to the best resolution strategy\n- Updates memory after execution so the agent never repeats the same mistake\n```\n\n**Use Case 2 — Expert Knowledge Distillation (SQL Copilot)**\n\n```python\nGoal: Help junior analysts solve tasks by reusing expert-level queries, patterns, and reasoning.\n\nUser: \"How do I calculate customer retention for this dataset?\"\n\nCognee tracks: expert SQL queries, workflow patterns, schema structures, successful implementations\n\n# Agent response:\nAgent: \"Here's how senior analysts solved a similar retention query.\n        Cognee matched your schema to a known structure and adapted\n        the expert's logic to fit your dataset.\"\n\n# What happens under the hood:\n- Extracts and stores patterns from expert SQL queries and workflows\n- Maps the current schema to previously seen structures\n- Retrieves similar tasks and their successful implementations\n- Adapts expert reasoning to the current context\n- Updates memory with new successful patterns so junior analysts perform at near-expert level\n```\n\n## Deploy Cognee\n\nUse [Cognee Cloud](https:\u002F\u002Fwww.cognee.ai) for a fully managed experience, or self-host with one of the 1-click deployment configurations below.\n\n| Platform | Best For | Command |\n|----------|----------|---------|\n| **Cognee Cloud** | Managed service, no infrastructure to maintain | [Sign up](https:\u002F\u002Fwww.cognee.ai) or `await cognee.serve()` |\n| **Modal** | Serverless, auto-scaling, GPU workloads | `bash distributed\u002Fdeploy\u002Fmodal-deploy.sh` |\n| **Railway** | Simplest PaaS, native Postgres | `railway init && railway up` |\n| **Fly.io** | Edge deployment, persistent volumes | `bash distributed\u002Fdeploy\u002Ffly-deploy.sh` |\n| **Render** | Simple PaaS with managed Postgres | Deploy to Render button |\n| **Daytona** | Cloud sandboxes (SDK or CLI) | See `distributed\u002Fdeploy\u002Fdaytona_sandbox.py` |\n\nSee the [`distributed\u002F`](distributed\u002F) folder for deploy scripts, worker configurations, and additional details.\n\n## Latest News\n\n[![Watch Demo](https:\u002F\u002Fimg.youtube.com\u002Fvi\u002F8hmqS2Y5RVQ\u002Fmaxresdefault.jpg)](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=8hmqS2Y5RVQ&t=13s)\n\n\n## Community & Support\n\n### Contributing\nWe welcome contributions from the community! Your input helps make Cognee better for everyone. See [`CONTRIBUTING.md`](CONTRIBUTING.md) to get started.\n\n### Code of Conduct\n\nWe're committed to fostering an inclusive and respectful community. Read our [Code of Conduct](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee\u002Fblob\u002Fmain\u002FCODE_OF_CONDUCT.md) for guidelines.\n\n## Research & Citation\n\nWe recently published a research paper on optimizing knowledge graphs for LLM reasoning:\n\n```bibtex\n@misc{markovic2025optimizinginterfaceknowledgegraphs,\n      title={Optimizing the Interface Between Knowledge Graphs and LLMs for Complex Reasoning},\n      author={Vasilije Markovic and Lazar Obradovic and Laszlo Hajdu and Jovan Pavlovic},\n      year={2025},\n      eprint={2505.24478},\n      archivePrefix={arXiv},\n      primaryClass={cs.AI},\n      url={https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.24478},\n}\n```\n","Cognee 是一个为AI代理提供共享且不断优化的记忆控制平面的项目，仅需6行代码即可实现。其核心功能包括通过图数据库和向量数据库技术构建认知记忆，使AI代理能够基于上下文进行回忆、连接并采取行动。该工具特别适合需要增强AI系统在数据处理、决策支持及工作流管理方面能力的应用场景，如智能客服、自动化办公流程等。采用Python语言开发，并遵循Apache License 2.0开源协议，鼓励社区贡献与协作。",2,"2026-06-11 02:50:41","top_language"]