patrick-toulme/harnessgym

Iterative agent harness improvement: run a coding agent on a hard task, generate the reusable tooling it was missing, qualify it, and replay fresh sessions with it activated. Works with Codex and Claude Code.

语言：Python · GitHub 星标：48 · Fork：6

话题：agents · ai-agents · benchmarking · claude-code · codex · coding-agents · developer-tools · harness

在 GitHub 上查看原仓库