{"skill":{"slug":"agent-research-harness","displayName":"research-harness","summary":"Cognitive discipline for AI-native scientific experimentation. Trigger when setting up controlled experiments with LLM agents, designing reproducible evaluat...","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":61,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1778508047886,"updatedAt":1778508110245},"latestVersion":{"version":"1.0.0","createdAt":1778508047886,"changelog":"v1.0.0 — 2026-05-11\n\n- Initial release distilled from real-world AI-native research execution practice\n- Five reference modules: repo architecture, experiment design, scoring/statistics, scientific thinking, agent collaboration\n- Core workflow: Phase 0 Scaffold → Phase 1 Harden → Phase 2 Design → Phase 3 Execute → Phase 4 Handoff\n- Six non-negotiables and five cognitive disciplines for agent-led research\n- Domain-agnostic: applicable to any field using LLM-based controlled experiments\n- SKILL.md as lean index; detailed content in references/ (no duplication)\n- Clean frontmatter with agent_created flag for proper skill lifecycle management","license":"MIT-0"},"metadata":null,"owner":{"handle":"zhelunsun","userId":"s178q41vf3nvhgdhvwtgx794p5858ce6","displayName":"Julian Zhelun Sun","image":"https://avatars.githubusercontent.com/u/202214845?v=4"},"moderation":null}