Tame Your Agents!
summary
Benoît Fontaine presents how to turn assistants like Claude or Cursor into reliable engineering partners by standardizing context using the AGENTS.md format. He explains the components of an effective AGENTS.md (personas, command limits, etc.), how to write User Stories that force agents to produce tests before code, and strategies for managing agent memory so the assistant follows architecture standards and business requirements.
AI-written summary · about
related talks
40mHow To Tame Your Agents?
24mFrom Prompts to AGENTS.md: What Survives Across Thousands of Runs
27mAgentic Engineering: Working With AI, Not Just Using It
49mAgile for Agents Managing Robots The Way We Manage Humans
27mAgentic Coding with Discipline & Skill: Practical Techniques and Tips
49mAI Agents: Collaboration in Real-Life Applications
19mBeyond agents: AI as a creative partner
16mBeyond the IDE: The Future of Developing Software With Agents