writing-skills
Use when creating, editing, or validating reusable agent skills and their trigger behavior. Do NOT use for ordinary project documentation or one-off prompts.
- License 许可
- MIT
- 最近更新
- 2026-10-14
- Stars
- ★ 16
- Forks
- 5
01 前言 Front Matter
- name
- writing-skills
- version
- 6.2.0
- category
- workflow
- description
- Use when creating, editing, or validating reusable agent skills and their trigger behavior. Do NOT use for ordinary project documentation or one-off prompts.
02 资源文件 7
- 参考文档anthropic-best-practices.md45.1 KB
- 参考文档CHANGELOG.md311 B
- 参考文档examples/CLAUDE_MD_TESTING.md5.3 KB
- 其他graphviz-conventions.dot5.8 KB
- 参考文档persuasion-principles.md5.8 KB
- 脚本render-graphs.js4.7 KB
- 参考文档testing-skills-with-subagents.md12.3 KB
03 SKILL.md 正文
Writing Skills
Treat skills as behavior-shaping interfaces. Validate them in proportion to the cost of a wrong instruction.
Read ../using-superpowers/references/risk-policy.md.
Authoring
- Keep
SKILL.mdfocused on a reusable trigger and workflow. - Frontmatter needs
name, a semverversion,category, and a description with explicitUse whenandDo NOT useconditions. - The description states triggering conditions, not the entire workflow.
- Prefer explicit conditions and positive recipes over long prohibition lists.
- Put heavy reference material or reusable scripts in supporting files.
- Remove duplicated guidance and stale examples.
- Frequently loaded skills should be short.
Validation Budget
| Change | Validation |
|---|---|
| Typo, link, metadata, wording clarification | Static validation and direct inspection |
| Ordinary workflow change | Static validation plus one focused scenario |
| Safety-critical discipline or broad automatic behavior | Baseline scenario, changed scenario, and additional cases only for observed loopholes |
A scenario may be an isolated manual prompt, a harness check, or a subagent when delegation is available and justified. Subagents, 5-repetition micro-tests, and adversarial review loops are not mandatory for every edit.
Do not create tests when the control does not exhibit the failure. Do not keep adding scenarios after the intended behavior is demonstrated unless a specific gap remains.
Checklist
- Confirm the trigger is narrow and searchable.
- Check instructions for contradictions and unbounded loops.
- Run the repository's skill or plugin validator.
- Run the risk-appropriate focused scenario.
- Inspect the final diff.
- Deploy through the platform's documented plugin update flow.
Do not commit or publish without user authorization.
04 同仓库更多 Skills
bad
暂无简介
writing-plans
Use when an approved change spans multiple dependent outcomes and needs a durable implementation sequence before coding. Do NOT use for small, obvious, reversible edits.
verification-before-completion
Use when about to claim work is complete, fixed, passing, or ready to integrate. Do NOT use to repeat checks without a completion claim or concrete doubt.
using-superpowers
Use when starting a conversation or deciding whether a Superpowers workflow matches the current task. Do NOT use it to auto-invoke skills whose stated triggers do not match.
using-git-worktrees
Use when feature work needs isolation from the current workspace or before executing a risky implementation plan. Do NOT use when the current workspace is already safe and isolated.
test-driven-development
Use when implementing a high-risk behavior change, reproduced regression, or costly contract. Do NOT use by default for docs, styling, config, or low-risk wiring.