Archived v0.1: what belongs in this catalog

[MicroMilo]

v0.1 is live with 19 reviewed papers on Claude Code and Codex CLI as complete production agents.\n\nThe catalog is built around one question: when a paper evaluates or outperforms a production coding agent, what changed, what improved, and was the comparison controlled?\n\nWe especially welcome:\n\n- missing papers that actually execute Claude Code or Codex CLI—not just a Claude or GPT-family model;\n- exact CLI version, model, reasoning mode, token, time, and cost evidence;\n- corrections with table, figure, section, or appendix locations;\n- official code, data, traces, and reproduction artifacts.\n\nStart with the paper-suggestion or correction issue template. Direct comparison does not mean fair comparison, and this repository is not a leaderboard: unknown controls stay explicitly unknown.\n\nNext after the v0.1 foundation: deeper evidence audits, venue-proceedings ingestion, and carefully scoped expansion to other industrial coding agents.