Awesome Tech Writing Skill Maintainer
Use this skill to maintain the skills that maintain this repository.
Load First
Read references/skill-design-notes.md before changing any file under
skills/.
Skill Maintenance Workflow
- Identify which maintenance behavior is changing:
- README curation.
- Contribution review.
- Skill design and validation.
- Deterministic checks or scripts.
- Read the affected skill's
SKILL.md, any referenced files, andagents/openai.yaml. - Keep
SKILL.mdfocused on instructions needed on every run. Move detailed rubrics, source notes, or longer checklists toreferences/. - Use imperative procedures, not generic declarations. Prefer exact commands, output templates, and validation gates.
- Keep descriptions trigger-oriented. The frontmatter
descriptionmust say what the skill does and when to use it. - Update
agents/openai.yamlif the skill name, purpose, or default prompt changes. - Validate every changed skill:
python3 "$CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py" skills/<skill-name>
- If the README validator changed, run it against the current README:
python3 skills/awesome-tech-writing-curator/scripts/validate_readme.py readme.md
Skill Design Standards
- Ground instructions in this repo's files and actual maintenance workflows.
- Include gotchas that prevent likely mistakes.
- Use progressive disclosure: link reference files by name and explain when to load them.
- Bundle scripts only for checks that should be deterministic.
- Do not copy external skill content wholesale. Extract reusable patterns and adapt them to this repo.
- Keep the skill suite small. Add a new skill only for a coherent workflow that would otherwise make an existing skill too broad.
Gotchas
- These skills live in the repository, not in the user's global Codex skills directory.
- This repo intentionally has minimal project structure. Do not introduce build systems or package managers just to validate Markdown.
- A skill can reference another repo-local skill's reference file by relative path, but avoid deep reference chains.