sqlite AGENTS.md

2026-05-27 · Source: Simon Willison's Weblog · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Intermediate, quick

Summary

SQLite recently introduced an AGENTS.md file, dated May 27, 2026, to guide interactions for AI agents targeting its codebase. The policy explicitly states SQLite does not accept agentic code contributions or pull requests without prior agreement, though human developers will review concise proof-of-concept pull requests. It does, however, welcome agentic bug reports that include reproducible test cases. A recent commit strengthened this stance by removing "(currently)" from the agentic code rejection. Concurrently, due to a flood of varying quality AI-generated bug reports, SQLite established a new, dedicated SQLite Bug Forum to manage these submissions, with D. Richard Hipp actively resolving issues.

Key takeaway

For AI Engineers or developers considering contributing to SQLite using automated agents, understand that the project has a firm policy against accepting agentic code. While your agent-generated bug reports with reproducible test cases are welcomed in the new SQLite Bug Forum, direct code contributions from agents are not. Focus your efforts on high-quality bug identification and adhere strictly to the AGENTS.md guidelines.

Key insights

SQLite formalizes strict policies for AI agent interactions, rejecting agentic code but accepting agent-generated bug reports.

Principles

• SQLite rejects agentic code contributions.
• Human review for concise PoC pull requests.
• Agentic bug reports with reproducible test cases are accepted.

In practice

• Submit agentic bug reports to the new forum.
• Do not submit agentic code as pull requests.
• Ensure PoC pull requests are concise.

Topics

• SQLite
• AI Agents
• Open-Source Policy
• Code Contributions
• Bug Reporting
• AGENTS.md

Code references

• sqlite/sqlite

Best for: CTO, VP of Engineering/Data, AI Architect, Software Engineer, AI Engineer, Director of AI/ML

Related on AIssential

• Writing a Good AGENTS.md — Effective `AGENTS.md` files are concise, human-written, and focus on essential, non-redundant information.
• Quality and Security Signals in AI-Generated Python Refactoring Pull Requests — AI-generated code refactoring shows mixed quality and security outcomes despite high developer acceptance.
• Orchard: An Open-Source Agentic Modeling Framework — Orchard provides an open-source framework for scalable agentic modeling, enabling reusable data, training, and evaluation across diverse domains.
• Oracle launches agentic AI tools for enterprise data management — Oracle's expanded AI Database integrates agentic AI directly with enterprise data, enhancing security, privacy, and performance.
• Benchmarking AI Agents on Kubernetes — AI coding agents struggle with system-wide impact analysis, often providing incomplete fixes despite effective code retrieval.
• Agents should interview you — AI agents can significantly accelerate development by interviewing users for requirements and automating code generation and testing.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.