{"full_name":"topoteretes/cognee","owner":"topoteretes","name":"cognee","shard":"1","url":"https://github.com/topoteretes/cognee","description":"Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory with small models for free","homepage":"https://www.cognee.ai","topics":["ai","cognitive-architecture","vector-database","ai-agents","graph-database","ai-memory","cognitive-memory","knowledge","knowledge-graph","graph-rag","contributions-welcome","good-first-issue","good-first-pr","help-wanted","open-source"],"tag_pages":["ai","ai-agents","ai-memory","contributions-welcome","good-first-issue","graph-database","help-wanted","knowledge","knowledge-graph","open-source","vector-database"],"language":"Python","license":"Apache-2.0","avatar":"https://avatars.githubusercontent.com/u/125468716?v=4","created_at":"2023-08-16T16:16:33Z","pushed_at":"2026-11-15T18:20:52Z","metrics":{"stars":33331,"forks":3518,"watchers":101,"issues":1770,"prs":4692,"releases":160,"issues_open":770,"issues_closed":1000,"prs_open":356,"prs_closed":1481,"prs_merged":2855,"discussions":147},"growth":{"day":{"stars":21,"forks":6,"base_day":"2026-11-15"},"week":{"stars":302,"forks":63,"base_day":"2026-11-04"}},"history":{"dates":["2026-10-28","2026-10-29","2026-10-30","2026-10-31","2026-11-01","2026-11-02","2026-11-03","2026-11-04","2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[32751,32780,32821,32863,32898,32933,33011,33029,33201,33243,33268,33282,33295,33310,33331],"forks":[3415,3422,3430,3436,3438,3438,3448,3455,3494,3502,3504,3505,3511,3512,3518],"watchers":[98,98,98,98,99,99,99,99,null,101,101,101,101,101,101],"issues_open":[null,null,null,null,null,null,null,null,null,772,770,773,771,770,770],"prs_open":[null,null,null,null,null,null,null,null,null,347,350,346,353,354,356]},"boards":{},"same_owner":[],"issues":[{"number":4778,"title":"Hackathon [Feature]: Add Rollbar data-source connector","comments":30,"created_at":"2026-08-28T09:50:54Z"},{"number":4811,"title":"Hackathon [Feature]: Add Reddit data-source connector","comments":29,"created_at":"2026-08-28T09:53:13Z"},{"number":3625,"title":"Hackathon: JSON Schema → Pydantic → cognee graph utilities — extend the base + full tests","comments":29,"created_at":"2026-06-28T18:31:58Z"},{"number":5776,"title":"Hackathon [Feature]: Add WooCommerce data-source connector","comments":28,"created_at":"2026-10-13T17:08:29Z"},{"number":5470,"title":"Hackathon [Feature]: Add Mattermost data-source connector","comments":28,"created_at":"2026-10-07T07:12:31Z"},{"number":4784,"title":"Hackathon [Feature]: Add Mailchimp data-source connector","comments":27,"created_at":"2026-08-28T09:51:19Z"},{"number":4723,"title":"Hackathon [Feature]: Add ReadMe data-source connector","comments":27,"created_at":"2026-08-28T09:47:05Z"},{"number":5466,"title":"Hackathon [Feature]: Add Backstage data-source connector","comments":26,"created_at":"2026-10-07T06:30:27Z"}],"discussions":[{"number":6722,"title":"Chopper: a small open model that cuts only the needed sentences out of retrieved text","comments":0,"category":"Show and tell","created_at":"2026-11-15T09:07:57Z"},{"number":6501,"title":"Clarification on consent for AML evaluation and leaderboard publication","comments":0,"category":"General","created_at":"2026-11-02T08:23:40Z"},{"number":6500,"title":"I built Cloud Alter Ego: an MCP connector that gives every AI chat your rules and memory as plain Markdown files in your own cloud storage","comments":0,"category":"Show and tell","created_at":"2026-11-02T08:21:16Z"},{"number":6376,"title":"Mergetober: could someone review integration proposal #6038?","comments":0,"category":"Q&A","created_at":"2026-10-27T16:41:22Z"},{"number":6296,"title":"Accountable Contributions pilot — Slid Phi Labs","comments":0,"category":"General","created_at":"2026-10-23T17:40:56Z"},{"number":6203,"title":"My GafBrain: Making All My AI Conversation History Accessible","comments":0,"category":"Show and tell","created_at":"2026-10-21T14:42:17Z"},{"number":6000,"title":"Cognee CLI included in AgentManager's tool catalog","comments":0,"category":"Show and tell","created_at":"2026-10-18T11:50:38Z"},{"number":5986,"title":"It's too slow","comments":3,"category":"General","created_at":"2026-10-18T00:59:58Z"}],"latest_release":{"name":"v1.6.8 \u0014 Integrations & Connectors","tag":"v1.6.8","published_at":"2026-11-12T22:31:22Z","body":"# v1.6.8 \u0014 Integrations & Connectors\n\n**Release Date:** 2026-11-12\n**Changes:** v1.6.7 → v1.6.8\n\n**Pull Requests:** #5189, #6410, #6417, #6593, #6612, #6634, #6642\n\n---\n\n## Summary\n\nThis release focuses on expanding and hardening integrations (new Slack support and many connector improvements) and on building reliable background work for embeddings and data updates. You’ll find a new Slack connector and app, better Google/Drive/Gmail/Calendar connector behavior, a RabbitMQ-based embedding worker option, and many reliability and provenance (audit) fixes so your imports, updates, and deletions behave more predictably.\n\n## Highlights\n\n- Added a fully-featured Slack connector and @cognee Slack app so Slack messages can be asked and remembered directly from Slack.\n- Introduced an embedding-worker engine over RabbitMQ so background embedding generation is more robust and scalable.\n- Improved Google-family connectors (Drive, Gmail, Calendar) including handling legacy file formats and retry logic so fewer files are skipped.\n- Many reliability fixes for updates and provenance: safer file locking, concurrent safe removals, and more correct audit (ledger) writes to avoid lost or out-of-order history.\n\n## New Features\n\n- Slack connector and @cognee app: A new integration that connects your Slack workspace to Cognee, letting users ask (search) and remember (save) information directly from Slack messages. Why it matters: you can interact with your Cognee memory without leaving Slack—ask questions, save snippets, and get answers where your team already communicates.\n- Embedding worker engine over RabbitMQ: Added a new embedding-worker backend that runs over RabbitMQ (a message broker) to generate vector embeddings (numeric representations used for semantic search) in the background. What it does: decouples embedding generation from the main service so it can scale independently and recover from worker failures. Why it matters: large or bursty ingestion workloads won’t block the main service, and embedding work can be retried and distributed more reliably.\n- Google Drive connector: Better ingestion of legacy Office formats (old .xls and .ppt files) and fixes for files that an earlier sync missed. What it does: reads and imports more types of Drive files and reconsiders files that previously failed. Why it matters: fewer missed documents and more complete content indexing from Google Drive.\n- Gmail and Calendar connector improvements: Updated source code and retry logic to make Gmail and Google Calendar ingestion more robust. What it does: improves how we call Google APIs and handle transient failures. Why it matters: more reliable imports from your email and calendar without manual re-runs.\n- POST /add accepts dlt_config: The add API now accepts a dlt_config parameter on POST /add to pass incremental sync options during ingestion. What it does: allows callers to configure the incremental loader when adding data. Why it matters: gives more control over how files are discovered and processed during ingestion.\n- Agent tool-output limit as constructor argument: The agent retriever now accepts the tool-output character limit as a constructor argument. What it does: makes it possible to configure how much text an agent tool is allowed to return. Why it matters: you can tune agents to avoid overly long outputs or to permit longer answers when needed.\n\n## Improvements\n\n- More robust update/removal flow: update tasks now remove released (deleted) files concurrently and in batches to speed cleanup and reduce lock contention. Why it matters: updates complete faster and storage cleanup is less likely to leave stale files behind.\n- Stored-file locking per location: Fixed and hardened how stored files are locked by location during updates so operations on one storage location don’t block others. Why it matters: reduces accidental contention when multiple update tasks touch different storage backends.\n- Provenance / ledger accuracy: Multiple fixes ensure th","is_prerelease":false},"readme":true,"readme_zh":false,"news":[{"title":"Show HN: Cognee – Open-Source AI Memory Layer That Remembers Context","url":"https://github.com/topoteretes/cognee","hn":"https://news.ycombinator.com/item?id=44169594","source":"Hacker News","lang":"en","date":"2025-06-03","points":9,"comments":2},{"title":"Show HN: Auto-optimizing deterministic LLM outputs using knowledge graphs","url":"https://github.com/topoteretes/cognee","hn":"https://news.ycombinator.com/item?id=40117217","source":"Hacker News","lang":"en","date":"2024-04-22","points":7,"comments":0},{"title":"Show HN: Cognee – Turn RAG and GraphRAG into custom dynamic semantic memory","url":"https://github.com/topoteretes/cognee","hn":"https://news.ycombinator.com/item?id=43031915","source":"Hacker News","lang":"en","date":"2025-02-13","points":6,"comments":1},{"title":"Show HN: I got Claude Code to run in Binary","url":"https://github.com/topoteretes/cognee-claude-code-binary","hn":"https://news.ycombinator.com/item?id=47616741","source":"Hacker News","lang":"en","date":"2026-04-02","points":2,"comments":0},{"title":"Cognee: Reliable LLM Memory for AI Applications and AI Agents","url":"https://github.com/topoteretes/cognee","hn":"https://news.ycombinator.com/item?id=42682745","source":"Hacker News","lang":"en","date":"2025-01-13","points":2,"comments":0},{"title":"Memory for AI agents in 6 lines of code","url":"https://github.com/topoteretes/cognee","hn":"https://news.ycombinator.com/item?id=46909419","source":"Hacker News","lang":"en","date":"2026-02-06","points":1,"comments":0},{"title":"终极部署指南：5步快速搭建AI记忆系统Cognee","url":"https://blog.csdn.net/gitblog_00333/article/details/158954140","source":"blog.csdn","lang":"zh","date":"2026-07-30"},{"title":"The Complete Guide to Implementing Persistent Memory in Hermes Agent with Cognee — From Knowledge Graph AI Memory Setup to Operation","url":"https://note.com/deft_gecko6923/n/nc0168676aa22?hl=en","source":"note","lang":"en","date":"2026-07-06"},{"title":"What is cognee, the tool that gives AI memory across sessions? A basic guide to running it via self-hosting","url":"https://note.com/shali_note/n/na5a186381736?hl=en","source":"note","lang":"en","date":"2026-07-05"},{"title":"cognee，有望替代 RAG, 简单了解一下","url":"https://blog.csdn.net/waterHBO/article/details/148513616","source":"blog.csdn","lang":"zh","date":"2025-06-08"},{"title":"Meet Lenny’s Memory: Building context graphs for AI agents","url":"https://news.google.com/rss/articles/CBMilAFBVV95cUxQZmdBb1lLWGROMUdKVDlmLTVKRnlwNnR1clRrazU1a2hCMkhhVWtYX1BMelZkdG1WY3M5ZXpldGZObkN6R09KWWZPZTMxY2RielBmb01tbWUtZUdEN3A2cmJIQ0Q1eEhxZjFYbFI5WkZZSWNrVFhZTEx6ZlhJVUo0bk0xUndlVWhjUk1DOEhtOHpKTVVz?oc=5","source":"Neo4j","lang":"en","date":"2026-02-02"}],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}