A curated list of resources for making AI agents truly understand you through Context Engineering.
Context Engineering is the practice of systematically providing AI with information about yourself—your preferences, documents, history, and workflows—to transform a generic AI assistant into one that truly understands you.
Every time you start a conversation with AI, it knows nothing about you:
- It doesn't know your coding style or tech stack
- It doesn't remember what you discussed yesterday
- It can't access your notes, documents, or data
- It treats you the same as millions of other users
Context Engineering solves this by injecting your personal context into AI interactions:
Your Context (what makes AI understand YOU)
├── Instructions: "I prefer TypeScript, be concise"
├── Background: Your role, expertise, current projects
├── Documents: Your notes, files, code
├── Memory: Facts AI learned about you over time
└── Connected Apps: Your Notion, Slack, email, calendar
- Effective Context Engineering for AI Agents - Anthropic's engineering blog
- Context Engineering in Agents: Memory Patterns - Memory patterns guide
- Building Personal Context - Reddit discussion
- Level 1: Platform Settings
- Level 2: Connecting Your Data
- Level 3: Personal Memory
- Level 4: Personal AI Stack
- All-in-One Solution
- Papers
- Resources
Basic personalization using built-in features. Takes 5 minutes.
- Custom Instructions Guide - Official guide for personalization
- Memory Feature FAQ - How ChatGPT remembers you
- Custom GPTs - Build personalized GPT
- Awesome ChatGPT Prompts - 1000+ prompt templates
- ChatGPT AutoExpert - Custom instructions for better responses
- Mr. Ranedeer AI Tutor - Personalized learning GPT
- Claude Projects - Project-based context
- Custom Instructions - Personalization features
- Claude Prompt Library - Official prompt examples
- Anthropic Cookbook - Official examples and recipes
- Claude Engineer - Engineering assistant setup
- Rules for AI - Official documentation
- Context Guide - Understanding cursor context
- cursor.directory - 1000+ community cursor rules
- Awesome CursorRules - Curated rules collection
- Cursor Rules by Pontus - Source for cursor.directory
- Cursor Rules Collection - Another curated collection
- Copilot Customization - Official docs
- Custom Instructions - Adding instructions
- Gemini Extensions - Connect Google services
- Gems (Custom Gemini) - Create personalized Gemini
Give AI access to your files, notes, and apps.
- MCP Documentation - Official docs
- MCP Announcement - Anthropic's blog
- Official MCP Servers - Reference implementations
- Awesome MCP Servers - Community servers list (800+)
- MCP Hub - Another server collection
- MCP Quickstart - Official setup guide
- Claude Desktop MCP Config - Configuration for Claude Desktop
Popular MCP Servers:
- Filesystem - Connect your local files
- Obsidian - Your Obsidian vault
- Obsidian Memory MCP - Memory as markdown in Obsidian
- Obsidian Palace - Enhanced Obsidian integration
- Notion - Your Notion pages
- Google Drive - Your Drive files
- Memory - Personal knowledge graph
- Sider - AI sidebar with page context
- Monica - Browser AI with memory
- Merlin - ChatGPT on any webpage
- MaxAI - Multi-model browser AI
- Zapier - Connect AI to 6000+ apps (Cloud)
- n8n - Open-source workflow automation (Self-hosted)
- Activepieces - Open-source Zapier alternative (Self-hosted)
AI that remembers you across conversations.
- ChatGPT Memory - Auto-learns from conversations
- Claude Projects - Persistent context per project
- Gemini Saved Info - Save information for Gemini
- Rewind - Records everything on Mac, AI-searchable
- Supermemory - Memory from bookmarks, tweets, articles
- Pieces - Long-term memory across dev tools
- Personal AI - AI that learns your knowledge
- Granola - AI notepad for meetings
- Memspan - File-first personal memory, portable
- Jean Memory - Remote personal memory for Claude
- Sem-Mem - Local semantic memory system
- MemLayer - Persistent memory for local LLMs
- Basic Memory - Personal knowledge graph system
- OpenMemory - Rich memory with natural decay
- Memori - SQL-native personal memory
- rag-user-memories - Extract & store personal facts
- Notion AI - AI over your Notion workspace
- Mem - AI-powered notes
- Reflect - AI-native note-taking
- Obsidian + AI - Local notes with AI plugins
- Reor - Local PKM with vector search
- Memos - Self-hosted, privacy-first notes
Self-hosted AI systems you fully control.
- Open WebUI - ChatGPT-like UI with memory & RAG
- Khoj - Personal AI for your notes/docs
- AnythingLLM - All-in-one document AI
- Quivr - Second brain with AI
- PrivateGPT - Fully private document chat
- Danswer - AI search over your data
- Perplexica - Personal AI search engine
- Second Me - AI self with hierarchical memory
- STING - Privacy-first desktop AI
- Tiiny AI - Local DB from your files
- Open WebUI Docker Setup - Official setup guide
- Khoj Self-Hosting - Self-hosting documentation
- Ollama - Run LLMs locally
- LM Studio - Desktop LLM app
- Jan - Offline AI assistant
- GPT4All - Local LLM ecosystem
- Llamafile - Single-file LLM distribution
- Chroma - Easy to start vector database
- LanceDB - Embedded, no server needed
- LangChain RAG Tutorial - Build personal RAG
- LlamaIndex Getting Started - Document AI concepts
- DocMind AI - Offline document analysis with GraphRAG
- Marker - PDF to Markdown
- Unstructured - Multi-format processing
Building your own personalization stack is powerful but complex. If you want everything above in one solution:
Membase - Universal Memory for your Agnets
A personal memory layer that gives you:
-
Cross-agent sync: Discuss plans with ChatGPT, then jump into Claude Code to build them without re-explaining context. Your memory follows you across all agents.
-
External context awareness: Automatically pulls context from your calendar, Notion, Slack, and emails. Your AI knows what's happening in your life without you explicitly saying it.
-
Token efficient: Instead of dumping everything into every session, Membase uses hybrid retrieval to inject only the relevant context for each request.
Foundational
- MemGPT: Towards LLMs as Operating Systems - Virtual context management with tiered memory
- Retrieval-Augmented Generation for Knowledge-Intensive NLP - The original RAG paper
Context Engineering
- Google's Context Engineering Whitepaper - Official Google whitepaper on sessions and memory
- Context Engineering 2.0 - Historical & conceptual grounding of context engineering
- Agentic Context Engineering (ACE) - Evolving contexts for self-improving LLMs
- Everything is Context - File system abstraction for context engineering
- Chain of Agents - Google's framework for long-context collaboration
- The Complexity Trap - Simple observation masking vs LLM summarization for context management
Personal Memory & Personalization
- USER-LLM - Google's user embedding for LLM personalization
- O-Mem: Omni Memory System - Hierarchical memory for personalized agents
- CUPID - Evaluating personalized alignment from interactions
- ComMer - Compressing user data for personalization
Memory Systems
- Memory in AI Agents: Taxonomies & Directions - Systematic memory classification
- A-MEM: Agentic Memory - Dynamic memory organization with Zettelkasten-style linking
- HiMeS - Hierarchical short + long-term memory
- Second Me - AI self with hierarchical memory
- Evo-Memory - Google's benchmarking of LLM agent test-time learning with self-evolving memory
Agent Systems & Orchestration
- Multi-Agent Collaboration via Evolving Orchestration - Evolving multi-agent collaboration patterns
- CodeAct: Executable Code Actions Elicit Better LLM Agents - Code-based agent actions
Evaluation & Benchmarks
- LOCCO - Evaluating long-term memory of LLMs
- MemoryAgentBench - Benchmarking memory via multi-turn interactions
- Context Engineering for AI Agents - Manus's lessons from building AI agents
- Context Rot - Chroma's research on context degradation
- Memory in AI Agents - Engineering long-term memory
- Agent Memory by Letta - Comprehensive memory guide
- Recursive Language Models - New paradigm for LLM architecture
- Awesome LLM - LLM resources
- Awesome LLM Apps - LLM applications
- Awesome AI Agents - AI agents list
- Awesome Local AI - Local AI resources
Contributions welcome! Please read CONTRIBUTING.md.
Looking for:
- Real examples and templates people have created
- Personal memory projects and services
- MCP servers and integrations
- Setup guides and tutorials
- Join our Discord - Chat with other developers personalizing agents
- Follow on Twitter - Stay updated on new updates and features
- Questions? support@aristo.so
Join 1,300+ developers personalizing agents
Apache License 2.0 - see LICENSE.