mgiovani/ai-arsenal / skills/create-skill
create-skill
VerifiedCreate a new agent skill (or Claude Code slash command) from a plain-language description, using live spec fetching, pattern research, and an approval-gated blueprint before any files are written. Use whenever the user wants to build, scaffold, or author a new skill, subagent capability, or slash command, including phrasings like 'make a command for X', 'create a slash command', 'turn this into a reusable skill', or 'package this workflow as a skill'. Not for editing CLAUDE.md/AGENTS.md memory rules (use create-rule) or discovering/installing existing third-party skills (use find-skills).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 20.3 KB |
| evals/evals.json | config | 3.5 KB |
| evals/trigger-eval.json | config | 2.2 KB |
| references/agent-prompts.md | markdown | 4.5 KB |
| references/eval-system.md | markdown | 5.6 KB |
| references/frontmatter-fields.md | markdown | 2.8 KB |
| references/schemas.md | markdown | 6.3 KB |
| references/skill-anatomy.md | markdown | 10.2 KB |
| references/specification-urls.md | markdown | 2.6 KB |
| scripts/generate_report.py | script | 11.5 KB |
| scripts/improve_description.py | script | 16.7 KB |
| scripts/package_skill.py | script | 4.5 KB |
| scripts/quick_validate.pyexecutable | script | 11.1 KB |
| scripts/run_eval.py | script | 8.1 KB |
evals/evals.json
View on GitHub{
"skill": "create-skill",
"description": "Creates new agent skills from a plain-language description using live spec fetching, pattern research, and an approval-gated blueprint",
"evals": [
{
"id": "happy-path-full-spec",
"prompt": "Create a skill called 'docs-glossary' that scans a docs/ directory for terms wrapped in **bold** on their first mention and generates a docs/glossary.md file listing each term with the sentence it was first defined in, sorted alphabetically. It should trigger on phrases like 'generate a glossary', 'build a term glossary from the docs', or 'list all the defined terms in docs/'. Success looks like a docs/glossary.md with one Markdown bullet per term. Yes, this is worth measuring — please include evals.",
"assertions": [
"Before any file under skills/ is created, the response presents a complete blueprint: proposed frontmatter (name: docs-glossary, a description under 1024 chars covering the stated trigger phrases), a directory structure with justification for each included/excluded subdir, and a workflow outline",
"The blueprint explicitly asks for or waits on approval (via plan-mode exit or an explicit question) rather than writing files immediately after presenting it",
"The response does not re-ask what the skill does, when it triggers, or what success looks like — that information was already given in the prompt and must be mined from it, not re-requested",
"The plan includes an evals/evals.json entry or explicitly states one will be created, since the prompt confirmed the skill is worth measuring",
"The proposed frontmatter does not include any key outside name, description, license, allowed-tools, metadata, compatibility, disable-model-invocation, argument-hint, context, agent, hooks"
]
},
{
"id": "judgment-overlapping-skill",
"prompt": "Make me a skill that reviews a pull request for security vulnerabilities like SQL injection and hardcoded secrets before merge.",
"assertions": [
"The response's research/composition step identifies that this repository already has a review-security skill covering OWASP-style security analysis, and surfaces that overlap to the user before proposing a brand-new skill",
"The Composition Plan section of the blueprint explicitly addresses reuse-vs-new-skill tradeoff (e.g. extend/invoke review-security, scope this differently, or justify why a new skill is still warranted) rather than silently blueprinting a duplicate with no mention of the existing skill",
"No skill files are written before the blueprint is presented and approval is requested"
]
},
{
"id": "edge-case-no-detail-refusal",
"prompt": "make me a skill",
"assertions": [
"The response does not fabricate a skill name, description, or purpose out of nothing — it asks the user what the skill should do, when it should trigger, and what success looks like (Phase 1's key questions), since none of that was supplied and there's no prior conversation to mine",
"The response contains no mkdir, Write, or Edit tool call creating any file or directory under skills/ or elsewhere",
"The response ends at the clarifying question(s) — it does not follow the question with a provisional guess or a partially-generated blueprint based on assumed details",
"No YAML frontmatter, directory tree, or workflow outline naming a specific invented skill is presented as if requirements were already known"
]
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |