{"id":28371781,"url":"https://github.com/jaffarkeikei/mynd","last_synced_at":"2026-04-29T09:34:47.214Z","repository":{"id":295269163,"uuid":"989559067","full_name":"jaffarkeikei/Mynd","owner":"jaffarkeikei","description":"Mynd is a universal memory layer for AI that automatically captures your digital context and streams it securely to any AI via Model Context Protocol (MCP). Your AIs finally remember everything about you - your decisions, preferences, history, and patterns - while your data never leaves your device.","archived":false,"fork":false,"pushed_at":"2025-05-24T22:11:25.000Z","size":158,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-06-22T02:40:36.415Z","etag":null,"topics":["chromadb","claude-api","gemini-api","mcp-server","python","shell","sqldb"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/jaffarkeikei.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"docs/SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-05-24T10:50:03.000Z","updated_at":"2025-05-28T10:25:08.000Z","dependencies_parsed_at":"2025-05-24T15:59:58.420Z","dependency_job_id":"d2a1f68d-210c-4b8b-b14e-052fc4c1d04b","html_url":"https://github.com/jaffarkeikei/Mynd","commit_stats":null,"previous_names":["jaffarkeikei/mynd"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/jaffarkeikei/Mynd","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jaffarkeikei%2FMynd","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jaffarkeikei%2FMynd/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jaffarkeikei%2FMynd/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jaffarkeikei%2FMynd/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jaffarkeikei","download_url":"https://codeload.github.com/jaffarkeikei/Mynd/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jaffarkeikei%2FMynd/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32420347,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-29T06:29:02.080Z","status":"ssl_error","status_checked_at":"2026-04-29T06:29:00.631Z","response_time":110,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["chromadb","claude-api","gemini-api","mcp-server","python","shell","sqldb"],"created_at":"2025-05-29T10:46:40.532Z","updated_at":"2026-04-29T09:34:47.208Z","avatar_url":"https://github.com/jaffarkeikei.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Mynd\n\n**_Give every AI a photographic memory of YOUR life - securely, locally, forever_**\n\n\u003e **The Problem:** Every AI conversation starts from zero. ChatGPT doesn't remember what you discussed yesterday. Copilot doesn't know your coding style. Claude forgets your preferences. It's like having digital Alzheimer's.\n\n\u003e **The Solution:** Mynd gives EVERY AI perfect memory of your context - securely, privately, forever.\n\n## What is Mynd?\n\nMynd is a **universal memory layer** for AI that automatically captures your digital context and streams it securely to any AI via Model Context Protocol (MCP). Your AIs finally remember everything about you - your decisions, preferences, history, and patterns - while your data never leaves your device.\n\n## 🌟 NEW: Beautiful Web Interface\n\nExperience Mynd through our **ChatGPT-like web interface** that visually demonstrates the power of AI memory:\n\n![Mynd Web Interface Demo](https://img.shields.io/badge/Live_Demo-Try_Now-blue?style=for-the-badge)\n\n```bash\n# Quick start the web demo\npython scripts/start_web_demo.py\n\n# Or manually:\npython src/web_app.py\n# Open http://localhost:8000\n```\n\n### Key Features:\n- **🎭 Memory Toggle** - Switch memory ON/OFF to see the dramatic difference\n- **🔄 Side-by-Side Comparison** - Compare responses with and without memory\n- **📊 Real-time Metrics** - Watch response time, relevance scores, and token usage\n- **💬 ChatGPT-like Interface** - Beautiful, familiar, and intuitive\n- **🎯 Demo Mode** - Pre-loaded context for instant demonstrations\n\n### See The Difference:\n1. Ask: \"What was our authentication decision?\"\n2. Toggle memory OFF and ask again\n3. Watch the AI go from \"I don't have context\" to perfect recall!\n\n## System Architecture\n\n### High-Level Architecture\n```mermaid\ngraph TB\n    subgraph \"Data Sources\"\n        Browser[\"🌐 Browser History\"]\n        Files[\"📄 Documents \u0026 Code\"]\n        Clipboard[\"📋 Clipboard\"]\n        Git[\"🔧 Git Repositories\"]\n    end\n    \n    subgraph \"Mynd Core\"\n        Capture[\"📥 Data Capture\"]\n        Extract[\"🧠 Semantic Extractor\"]\n        Privacy[\"🔒 Privacy Filter\"]\n        \n        subgraph \"Storage\"\n            SQLite[\"📊 SQLite DB\u003cbr/\u003e(Metadata)\"]\n            ChromaDB[\"🧠 ChromaDB\u003cbr/\u003e(Vectors)\"]\n        end\n        \n        MCP[\"🔗 MCP Server\"]\n    end\n    \n    subgraph \"AI Clients\"\n        ChatGPT[\"💬 ChatGPT\"]\n        Claude[\"🤖 Claude\"]\n        Copilot[\"👨‍💻 GitHub Copilot\"]\n        AnyAI[\"🤖 Any AI Tool\"]\n    end\n    \n    Browser --\u003e Capture\n    Files --\u003e Capture\n    Clipboard --\u003e Capture\n    Git --\u003e Capture\n    \n    Capture --\u003e Extract\n    Extract --\u003e Privacy\n    Privacy --\u003e SQLite\n    Privacy --\u003e ChromaDB\n    \n    SQLite --\u003e MCP\n    ChromaDB --\u003e MCP\n    \n    MCP --\u003e|\"Secure Context\"| ChatGPT\n    MCP --\u003e|\"Secure Context\"| Claude\n    MCP --\u003e|\"Secure Context\"| Copilot\n    MCP --\u003e|\"Secure Context\"| AnyAI\n    \n    style Extract fill:#ff6b6b,stroke:#fff,stroke-width:3px\n    style Privacy fill:#4ecdc4,stroke:#333,stroke-width:2px\n    style MCP fill:#f39c12,stroke:#333,stroke-width:2px\n```\n\n### Data Flow Process\n```mermaid\nsequenceDiagram\n    participant U as User Activity\n    participant C as Data Capture\n    participant E as Semantic Extractor\n    participant P as Privacy Filter\n    participant D as Database\n    participant V as Vector Store\n    participant M as MCP Server\n    participant A as AI Client\n    \n    U-\u003e\u003eC: Browser/File/Code Activity\n    C-\u003e\u003eE: Raw Content\n    E-\u003e\u003eE: Extract Semantic Meaning\n    E-\u003e\u003eP: Semantic Events\n    P-\u003e\u003eP: Remove PII \u0026 Sensitive Data\n    P-\u003e\u003eD: Store Metadata\n    P-\u003e\u003eV: Store Embeddings\n    \n    Note over D,V: Local Storage Only\n    \n    A-\u003e\u003eM: Request Context for Query\n    M-\u003e\u003eV: Semantic Search\n    M-\u003e\u003eD: Get Related Events\n    M-\u003e\u003eM: Compress \u0026 Optimize\n    M-\u003e\u003eA: Relevant Context (4000 tokens max)\n    \n    Note over M,A: MCP Protocol\n```\n\n## The Memory Crisis (The $2.3T Problem)\n\n**Every AI interaction wastes massive time on context setup:**\n\n- **73% of AI conversations** repeat information from previous chats\n- **2.3 hours daily** lost re-explaining context to AI\n- **$2.3 trillion annually** in global productivity loss\n- **89% of professionals** frustrated with AI's goldfish memory\n\n**Real Examples:**\n- \"What was that API decision we made last month?\" → *\"I don't have context\"*\n- \"Continue our React project\" → *\"Can you share the codebase?\"*\n- \"Remember my coding style preferences\" → *\"Please describe them again\"*\n\n## Mynd Demo Script (2 Minutes)\n\n```bash\n# The Setup (30 seconds)\n\"Every AI suffers from digital amnesia. Watch this...\"\n\n[User asks ChatGPT]: \"What was that authentication architecture decision from last month?\"\n[ChatGPT]: \"I don't have access to previous conversations...\"\n\n# The Magic (60 seconds)\n[Install Mynd]: mynd demo\n[Capture context]: \"Mynd has been learning your patterns...\"\n\n[Same question to ChatGPT + Mynd]:\nmynd query \"authentication architecture decision\"\n\n[Result]: \"You decided on JWT with refresh tokens over sessions on March 15th \nbecause of mobile app requirements. You were concerned about XSS attacks but \nchose client-side storage anyway because your team lacks Redis expertise.\"\n\n# The Jaw-Drop (30 seconds)\n\"This context came from:\n✅ Your browser research from 6 weeks ago\n✅ Code comments you wrote in March  \n✅ A design doc you saved locally\n✅ All delivered securely via MCP - your data never left your machine\"\n```\n\n## Quick Start (2 Minutes to Life-Changing AI)\n\n### Component Initialization Flow\n```mermaid\ngraph LR\n    subgraph \"Setup Process\"\n        Install[\"🔧 Install Dependencies\"]\n        Init[\"🎯 Initialize Components\"]\n        Demo[\"🎬 Create Demo Data\"]\n        Query[\"🔍 Test Query\"]\n    end\n    \n    Install --\u003e Init\n    Init --\u003e Demo\n    Demo --\u003e Query\n    \n    subgraph \"Components Initialized\"\n        DB[\"📊 SQLite Database\"]\n        Vector[\"🧠 Vector Store\"]\n        Extractor[\"🔍 Semantic Extractor\"]\n        CLI[\"💻 CLI Interface\"]\n    end\n    \n    Init --\u003e DB\n    Init --\u003e Vector\n    Init --\u003e Extractor\n    Init --\u003e CLI\n```\n\n```bash\n# Install Mynd\n./install.sh  # or pip install -e .\n\n# Set up demo data\nmynd demo\n\n# Test the magic\nmynd query \"authentication architecture\"\n\n# Watch AI get perfect memory of your decisions!\n```\n\n## AgentHacks 2025 Categories\n\n### **PRIMARY: Personalization \u0026 Memory** \n- ✅ **Learns from user activity**: Continuous semantic capture\n- ✅ **Evolves behavior over time**: Memory graph grows and improves\n- ✅ **User corrections improve system**: Feedback loop for better context\n- ✅ **Personal preference adaptation**: Learns your patterns and style\n\n### **SECONDARY: Interfaces for Human-AI Collaboration**\n- ✅ **Revolutionizes AI interaction**: No more context re-explanation\n- ✅ **Seamless collaboration**: AI knows your full background\n- ✅ **Natural communication**: AI understands your references and history\n\n## Business Model \u0026 Market\n\n### Market Size\n- **TAM**: $450B (Global productivity software market)\n- **SAM**: $67B (AI tools and services) \n- **SOM**: $12B (AI productivity and memory solutions)\n\n### Revenue Model\n```mermaid\ngraph TD\n    Personal[\"🆓 Mynd Personal\u003cbr/\u003eFREE Forever\u003cbr/\u003e• 30-day memory\u003cbr/\u003e• 3 data sources\u003cbr/\u003e• Community support\"] \n    \n    Pro[\"💎 Mynd Pro\u003cbr/\u003e$29/month\u003cbr/\u003e• Unlimited memory\u003cbr/\u003e• All data sources\u003cbr/\u003e• Priority MCP access\u003cbr/\u003e• Advanced privacy controls\"]\n    \n    Enterprise[\"🏢 Mynd Enterprise\u003cbr/\u003e$199/user/month\u003cbr/\u003e• Team memory sharing\u003cbr/\u003e• Compliance controls\u003cbr/\u003e• Custom integrations\u003cbr/\u003e• White-label deployment\"]\n    \n    Personal --\u003e Pro\n    Pro --\u003e Enterprise\n    \n    style Personal fill:#4ecdc4\n    style Pro fill:#f39c12\n    style Enterprise fill:#e74c3c\n```\n\n## Security \u0026 Privacy Architecture\n\n### Privacy-First Data Flow\n```mermaid\ngraph TB\n    subgraph \"Your Device (Secure Zone)\"\n        Raw[\"📝 Raw Data\u003cbr/\u003e(Browser, Files, Code)\"]\n        PII[\"🔒 PII Detection\u003cbr/\u003e(Remove Sensitive Info)\"]\n        LLM[\"🧠 Local LLM\u003cbr/\u003e(Semantic Extraction)\"]\n        Encrypt[\"🔐 Encrypted Storage\u003cbr/\u003e(SQLite + ChromaDB)\"]\n    end\n    \n    subgraph \"External AI (Untrusted)\"\n        ChatGPT[\"💬 ChatGPT\"]\n        Claude[\"🤖 Claude\"] \n        Other[\"🤖 Other AIs\"]\n    end\n    \n    Raw --\u003e PII\n    PII --\u003e LLM\n    LLM --\u003e Encrypt\n    \n    Encrypt --\u003e|\"Semantic Context Only\u003cbr/\u003e(No Raw Data)\"| ChatGPT\n    Encrypt --\u003e|\"Semantic Context Only\u003cbr/\u003e(No Raw Data)\"| Claude\n    Encrypt --\u003e|\"Semantic Context Only\u003cbr/\u003e(No Raw Data)\"| Other\n    \n    style Raw fill:#ff6b6b,stroke:#333,stroke-width:2px\n    style PII fill:#4ecdc4,stroke:#333,stroke-width:2px\n    style LLM fill:#f39c12,stroke:#333,stroke-width:2px\n    style Encrypt fill:#27ae60,stroke:#333,stroke-width:2px\n```\n\n**Privacy Promise**: Your raw data NEVER leaves your device. Only semantic meaning is processed, stored locally, and delivered via encrypted MCP.\n\n## Success Metrics \u0026 Validation\n\n### Technical Milestones ✅\n- [x] Core semantic extraction engine (Local LLM + privacy filters)\n- [x] Local encrypted storage (ChromaDB + SQLite)\n- [x] MCP server architecture with capability tokens\n- [x] Browser history and document capture framework\n- [x] CLI interface with full functionality\n\n### Demo Readiness ✅ \n- [x] 2-minute live demo script prepared\n- [x] Real context database with semantic events\n- [x] Multiple query examples working\n- [x] Clear before/after comparison ready\n\n## Join the Memory Revolution\n\nMynd isn't just a hackathon project - it's the future of AI interaction. We're building the memory layer that every AI desperately needs.\n\n**For Developers**: Finally, coding AI that knows your entire project history\n**For Knowledge Workers**: AI assistants that remember every decision and context  \n**For Everyone**: The end of explaining the same thing to AI over and over\n\n---","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjaffarkeikei%2Fmynd","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjaffarkeikei%2Fmynd","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjaffarkeikei%2Fmynd/lists"}