ETH Zurich Research Questions Effectiveness of AGENTS.md Files for AI Coding Agents
By
noemit
Hand-rolled, kettle-boiled, baked to perfection. Worth every minute at the bakery.
Summary
ETH Zurich researchers published a paper challenging the value of AGENTS.md files for AI coding agents, finding they often hinder rather than help. The study recommends omitting LLM-generated context files entirely and limiting human-written instructions to only non-inferable details like specific tooling or custom build commands. The research team analyzed 60,000 open-source repositories and found that while context files are widely recommended, they frequently contain redundant or counterproductive information for AI coding agents.
Key quotes
· 3 pulledDespite widespread industry recommendations, a new ETH Zurich paper concludes that AGENTS.md files may often hinder AI coding agents.
The researchers recommend omitting LLM-generated context files entirely and limiting human-written instructions to non-inferable details, such as highly specific tooling or custom build commands.
The team behind the work, including Thibaud Gloaguen, Niels Mündler, Mark Müller, Veselin Raychev, and Martin Vechev, justified the research by noting that while 60,000 open-source repositories currently contain context files such as AGENTS.md...
You might also wanna read
AGENTS.md: An Open Format for Guiding AI Coding Agents in Open-Source Projects
AGENTS.md is a simple, open format for guiding AI coding agents, functioning as a README specifically designed for agents rather than humans
agents.md·2d ago
AGENTS.md: Standardized Documentation Format for AI Agents Adopted by Major Platforms
The article introduces AGENTS.md, a standardized format for AI agents that serves as a structured alternative to human-readable README files
Product Hunt·9mo ago
How AI coding agents are reshaping social science research: Opportunities and concerns
This article examines how AI coding agents are transforming social science research by automating core research tasks traditionally performe
share.google·4d ago
5 Tips for Redesigning Assignments in the Age of Agentic AI
The article argues that higher education's default response to agentic AI tools (like Einstein, built on Claude Code and Codex) that can aut
insidehighered.com·2d ago
AI-First Content Management: Rethinking CMS vs Markdown for Agentic Applications
The article explores whether traditional Content Management Systems (CMS) like WordPress are still necessary in an AI-first world where agen
Prototypr·4mo ago
How I Used Coding Agents to Automate My AI Research Work in Copilot Applied Science
An AI researcher shares their experience using coding agents to automate intellectual work, specifically building agents that automate parts
GitHub·2mo ago