当发现既有 workflow/测试写法不正确或缺失时:自动补齐/新增 skill,并在满足门禁时自动提交并 push。
How this skill is triggered — by the user, by Claude, or both
Slash command
/polyglot-agent-harness:auto-skill-maintainerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
当你明确发现以下任一情况时,必须执行本技能:
当你明确发现以下任一情况时,必须执行本技能:
pr-workflow 或跨技能交付策略变更(例如 fmt 门禁顺序、手动 PR 提交、PR 草稿落点、上游正文规范、commit message 不含 Co-authored-by)— 必须在同一次 commit 中同步更新所有受影响的 skills(至少:experiment-guard、completion-examiner、report-generator、test-authoring、harness-overview),避免知识分叉注意:该技能描述的是“必须做什么”,不代表运行环境会自动触发或自动执行。若工具侧不支持自动 hook,请在交付时明确说明已按本技能更新并提交。
origin remoteHARNESS_AUTO_PUSH=1--pushpython3 scripts/auto-skill.py ensure --name <skill-name> --description '<desc>' --commitskills/<skill-name>/SKILL.md(覆盖占位内容),并再次提交:
git add skills/<skill-name>/SKILL.md plugin.jsongit commit -m 'harness: refine <skill-name>'python3 -m json.tool plugin.json >/dev/nullHARNESS_AUTO_PUSH=1 python3 scripts/auto-skill.py ensure --name <skill-name> --description '<desc>' --pushskills/<name>/SKILL.mdnpx claudepluginhub jiaxin2006/polyglot-agent-harness --plugin polyglot-agent-harnessProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.