version: "1.0.0" name: agent-skills-creator description: Guides creation and improvement of best-practice agent skills following the open format specification. Covers frontmatter, directory structure, progressive disclosure, reference files, rules folders, degrees of freedom, content patterns, executable scripts, MCP tool references, evaluations, cross-model testing, and a ten-dimension audit protocol for existing skills. Use when creating a new skill, authoring SKILL.md, setting up a rules-based audit skill, structuring a skill bundle, writing scripts inside a skill, evaluating a skill, improving or rewriting an existing skill, or asking "how to write a skill", "improve this skill", "audit my skill", or "review this SKILL.md".

Agent Skills Creator

Create and improve skills in the Agent Skills open format: full lifecycle from pattern selection through validation and README update.

• IS: creating new agent skills and auditing or rewriting existing ones: SKILL.md, references, rules folders, scripts, evaluations.
• IS NOT: AGENTS.md/CLAUDE.md instruction files (use agents-md) or general documentation quality (use docs-writing).

Choose a Mode

• New skill → Creation Workflow below.
• Audit, improve, or rewrite an existing skill → load references/improving-existing-skills.md (scores ten audit dimensions, runs an ordered rewrite, then reuses Steps 5-8 for validation and shipping).

Reference Files

Choose a Skill Category

Determine what problem the skill solves; category informs pattern choice.

Load references/skill-categories.md for per-category guidance, tips, and examples.

Choose a Skill Pattern

Decision guide:

• Auditing or linting against a checklist: rules-based
• Guiding a multi-step process: workflow
• Dispatching to different tracks by context: simple/hub
• Unsure: start with workflow (most flexible)

Load references/skill-patterns.md for each pattern's templates and skeletons.

Creation Workflow

Copy this checklist to track progress:

Skill creation progress:
- [ ] Step 1: Choose skill category and pattern
- [ ] Step 2: Create directory and frontmatter
- [ ] Step 3: Write SKILL.md body
- [ ] Step 4: Add reference or rule files
- [ ] Step 5: Validate with quality checklist
- [ ] Step 6: Update README.md
- [ ] Step 7: Smoke-test installation
- [ ] Step 8: Evaluate and iterate

Step 1: Choose skill category and pattern

Determine the category (what problem it solves), then the structural pattern. Load references/skill-categories.md and references/skill-patterns.md.

Step 2: Create directory and frontmatter

Load references/format-specification.md for hard constraints.

• Create skills/<name>/SKILL.md
• Folder name must match name field (kebab-case)
• name: max 64 chars, lowercase letters/numbers/hyphens, no "anthropic" or "claude"
• description: max 1024 chars, third-person voice, include "Use when..." triggers with specific keywords

Step 3: Write SKILL.md body

Load references/authoring-tips.md for voice, degrees of freedom, content patterns, and descriptions. Apply:

• Keep under 500 lines; split into reference files if longer
• Open with an IS/IS-NOT pair when adjacent skills exist or scope creep is likely ("Open with Boundaries")
• Add only context Claude lacks ("Don't State the Obvious"); use consistent terminology
• Match degrees of freedom to fragility: prose for open-ended work, specific scripts for fragile/destructive ops ("Degrees of Freedom")
• Reach for named content patterns: template for fixed output, examples for format-sensitive output, conditional for decision points
• Add a copyable progress checklist for multi-step workflows; validation/feedback loops for quality-critical tasks
• Build a Gotchas section from observed failures: the highest-signal content

Step 4: Add reference or rule files

• Workflow/mixed: references/ folder of focused files, each linked from SKILL.md via a "Read when..." table.
• Rules-based: rules/ folder; load references/rules-folder-structure.md for _sections.md, _template.md, file-naming, and priority-table layout.
• Simple/hub: track files alongside SKILL.md, linked from a tracks table.

Key constraints:

• References one level deep from SKILL.md (no chains); loaded only when explicitly listed
• Files over 100 lines need a table of contents
• Long files (up to ~450 lines) are fine when TOC'd and single-topic; split by loading condition, not line count
• Broad domains: prefer a comprehensive-reference folder of many small files over monoliths (see the comprehensive-reference variant in references/skill-patterns.md)
• agents/ is a sanctioned optional folder for subagent prompt definitions the skill dispatches to

Advanced:

• scripts/ for executable utilities Claude composes; load references/executable-code.md for error handling, constants, plan-validate-execute, runtime, package deps, MCP tool naming
• config.json for skills needing user-specific setup context across sessions
• On-demand hooks (PreToolUse/PostToolUse) for safety gates or observation

Step 5: Validate

Load references/quality-checklist.md and run all applicable checks. Copyable local validation block:

ruby -ryaml -e 'path=ARGV.fetch(0); s=File.read(path); m=s.match(/\A---\n(.*?)\n---\n/m) or abort("missing frontmatter"); y=YAML.safe_load(m[1]); abort("missing name") if y["name"].to_s.empty?; d=y["description"].to_s; abort("missing Use when") unless d.include?("Use when"); abort("description too long #{d.length}") if d.length>1024; abort("body too long #{s.lines.size}") if s.lines.size>=500; puts "OK #{path}"' skills/<name>/SKILL.md
wc -l skills/<name>/SKILL.md
perl -CSD -ne 'print "$ARGV:$.:$_" if /\x{2014}/' skills/<name>/SKILL.md README.md docs/skills.mdx
rg -n "references/|scripts/|assets/" skills/<name>/SKILL.md
find skills/<name> -maxdepth 2 -type f | sort
rg -n "^- \\*\\*\\[<name>\\]" README.md docs/skills.mdx

Step 6: Update README.md

Add a bullet under the matching category heading and bump the skill count near the top of the README:

-**[<skill-name>](./skills/<skill-name>/SKILL.md)**: <one-line description>

Categories used in this repo: Architecture, Design, Writing, Quality, Shipping, Authoring.

Step 7: Smoke-test

Install via the skills CLI, never cp -R into ~/.claude/skills/ (bypasses the ~/.agents/skills symlink chain):

npx skills add mblode/agent-skills -g --skill <name> -y
ls ~/.claude/skills/<name>/

Deploy chain: skills add writes to ~/.agents/skills/<name>/, symlinked into ~/.claude/skills/<name>/ for Claude Code to pick up.

For local iteration without reinstalling, symlink the repo folder directly; unlink when done:

ln -s /path/to/agent-skills/skills/<name> ~/.claude/skills/<name>

Step 8: Evaluate and iterate

Load references/evaluation-and-iteration.md. Define 3+ scenarios, test on each target model, and iterate on observed Claude behavior, not assumptions.

Anti-patterns

• Dumping the full specification into the SKILL.md body (use reference files)
• Reference-to-reference chains (keep one level deep)
• Time-sensitive content ("before August 2025, use...")
• Restating what Claude already knows (Markdown, general coding advice, standard conventions)
• Description written as a human summary or in "I audit..."/"Use this to..." voice instead of a third-person model trigger with "Use when..." phrases and quoted user phrases
• Adding README.md, CHANGELOG.md, or INSTALLATION_GUIDE.md to the skill folder
• Dropping files in folders without linking them from SKILL.md
• Installing with cp -R into ~/.claude/skills/ (bypasses the ~/.agents/skills symlink chain); use npx skills add
• Leaving supporting files at the skill root for a non-simple/hub skill (move them to references/)
• Over-constraining the approach when specifying outcomes would suffice (railroading)
• Enumerating many tool or library options instead of giving one sensible default with an escape hatch for the known exception
• Vague or generic skill names (helper, utils, tools, documents, data) that give the model nothing to route on
• Skipping a Gotchas section for skills with known failure modes
• Storing persistent data in the skill directory (deleted on upgrade) or via hardcoded absolute paths instead of ${CLAUDE_PLUGIN_DATA}
• Referencing MCP tools without the server prefix (bigquery_schema instead of BigQuery:bigquery_schema)
• Magic numbers in scripts with no justifying comment (voodoo constants)
• Shipping without testing on every target model; what reads well to Opus may underspecify for Haiku

Related Skills

• agents-md for auditing AGENTS.md/CLAUDE.md instruction files
• docs-writing for documentation quality rules