AI agents
How agents are built, tuned and run: loops, harnesses, MCP, skills and coding agents, with what broke and what we measured.
Best notes
- Five layers to tune before you fork your agent
- Dispatch #2: handoff pattern: từ ảnh chụp app đến Figma và video demo
- What the Engelberg report got right about us
- Claude Code 2.1.259 turned our Read() deny rules into a Bash tripwire
- The debug loop: make the agent measure the browser before it touches the CSS
Reports
Writing
- Five layers to tune before you fork your agent
- Dispatch #2: handoff pattern: từ ảnh chụp app đến Figma và video demo
- What the Engelberg report got right about us
- Claude Code 2.1.259 turned our Read() deny rules into a Bash tripwire
- The debug loop: make the agent measure the browser before it touches the CSS
- DeepSeek Harness, dissected
- CAP breakdown
- Stagehand breakdown
- Context7 breakdown
- E2b breakdown
Show all 21 pieces
Research notes
- The Coding Agent Team
- How we crafted the OGIF summarizer bot to streamline weekly knowledge-sharing
- How we created an AI powered interview system using OpenAI's ChatGPT
- How knowledge work organizes itself
- The five stages of learning
- Should AI agents transact without human confirmation?
- AI agent explained
- Vibe culture
Show all 18 notes
- Mapping the path to AGI
- Securing your remote MCP servers
- Tool-level security for remote MCP servers
- Human touch in agent systems
- Evaluate chatbot agent by user simulation
- Building agent supervisors to generate insights
- Build custom AI agent with ElizaOS
- Multi-agent collaboration for task completion
- Function calling in AI agents
- Explaining gradient descent in machine learning with a simple analogy
Series
Tags in this topic
architecture26, mcp15, security11, breakdown10, agents9, commentary5, agent4, automation4, agentic-ai3, discord3, integration3, ai-agents2
Writers
tieubao11, monotykamary10, chinhld123, minhluuquang3, nnhuyhoang3, zlatanpham3, hthai22012, innnotruong2