skill-creatorlisted

Author or revise an AI-agent skill. Use when the user says "let's make a skill for X", invokes /skill-creator, asks to revise an existing SKILL.md, or wants to skill-ify a repeated workflow with 2+ prior invocations. Inputs - trigger phrase, artifact, evidence tier (E empirical / P practitioner / H heuristic). Do not use when the workflow is one-shot (write a snippet), is a phrasing trick or persona without a stable model, is an eval/benchmark loop (separate kit), or duplicates an existing skill (merge). Produces a skill directory with SKILL.md (5-part trigger description, lean body, no README/INSTALL/CHANGELOG) plus the registry-integration step for the active backend (see references/registry-integration.md). Escalate if no repeated workflow exists (suggest snippet), the model is H-tier with overreach ≥4 (manual-only), or the skill would compete with a stronger existing one (merge).
Naoray/skills · ★ 1 · AI & Automation · score 64

Install: claude install-skill Naoray/skills

# Skill Creator Package a repeated workflow as a triggerable artifact so future-agents find and execute the procedure without re-deriving it. **Evidence tier**: P (practitioner-backed) **Basis**: Anthropic's `skill-creator`, OpenAI's `skill-creator`, and the *Principled Framework for Curating AI-Agent Skills* (5-part trigger contract, evidence tiers, kit/skill/snippet forms, rubric overrides), plus a two-pass counselor panel + ChatGPT Deep Research synthesis on multi-mode packaging, evals, lifecycle, and portability. **Source IDs**: anthropics/skills, openai/skills, Naoray/skills, *Corpus2Skill* and *Navigation Paradox* on explicit routing, *Block skill-design* on skills-as-code. **Reviewed**: 2026-05-12 Treat as operating procedure with sharp boundaries, not universal best practice. ## Fit gate — run BEFORE any file work Score the proposed skill. If 2+ are no, do NOT create it. Output the reason and the recommended alternative form (snippet, inline agent prompt, merge into existing skill, archive). 1. **Repeated workflow.** ≥2 prior real invocations exist. Not a one-shot. 2. **Stable artifact.** Same output shape across runs. 3. **Durable cognitive model.** Domain model, standard, expert workflow, or validated heuristic — not a phrasing trick or persona. 4. **Sharp trigger boundary.** Specific phrase or context, not "when the user wants something nice". 5. **Unique slot.** No existing skill in the registry covers it. Check folder + catalog (see `references/registry-int