Agent behavior
A standard format for describing expected agent behavior for trace review and evaluation. Behavior specs live in .agents/behaviors/ and define what good agent conduct looks like before a reviewer, rubric, scorer, or eval checks for it.
Define what good agent behavior looks like Agent behavior is a format for writing down the behavior you expect an AI agent to follow across many interactions. Each behavior spec is a Markdown file that lives in your repo and describes the recurring conduct that makes the agent reliable. The spec captures that standard up front, so reviewers, rubrics, scorers, and evals have something concrete to measure against. .agents/behaviors/ .agents/behaviors/ └── financial-work-verification/ └── BEHAVIOR.md Written for review Specs speak to the people and agents who read traces, design evals,…
saved by
related reading
- Agentic Evals Pyramidrwilinski.ai
- Is Agentic: AI Agent Readiness Score for your Site and Appis-agentic.com
- Senior SWE-Benchsenior-swe-bench.snorkel.ai
- Building Effective AI Agents \ Anthropicanthropic.com
- Demystifying evals for AI agents \ Anthropicanthropic.com
- Senior SWE-Benchsenior-swe-bench.snorkel.ai
- Building Effective AI Agents \ Anthropicanthropic.com
- AGENTS.mdagents.md
- Agent Evaluation: A Detailed Guidecameronrwolfe.substack.com
- AddyOsmani.com - How to write a good spec for AI agentsaddyosmani.com
- DESIGN.md Generator — Design System for AI Agents | design.devdesign.dev
- Best practices for coding with agents · Cursorcursor.com