Welcome: help build the evidence map for production coding agents

[MicroMilo]

This repository is now public. It tracks formally accepted 2026 conference papers that actually run or evaluate Claude Code or Codex CLI as products.

Start here:

• Interactive catalog: https://micromilo.github.io/awesome-claude-code-codex-papers/
• Evidence-backed insights: https://micromilo.github.io/awesome-claude-code-codex-papers/insights/
• Contribution rules: https://github.com/MicroMilo/awesome-claude-code-codex-papers/blob/main/CONTRIBUTING.md
• Full conference census: https://github.com/MicroMilo/awesome-claude-code-codex-papers/blob/main/docs/2026-conference-census.md

Use Discussions for research questions, scope debates, and ideas. Use Issues for concrete paper suggestions, metadata corrections, and implementation work. Every claim should remain traceable to a first-party source, and uncertainty should stay visible.

What failure mode, benchmark, or conference source should the catalog investigate next?