Skill Creator

Skill Creator

Creates complete, spec-compliant skill packages following AgentSkills.io and Anthropic standards. Supports both creation and validation workflows with 100-point marketplace grading.

Overview

Skill Creator solves the gap between writing ad-hoc agent skills and producing marketplace-ready packages that score well on the Intent Solutions 100-point rubric. It enforces the 2026 spec (top-level identity fields, ${CLAUDE_SKILL_DIR} paths, scored sections) and catches contradictions that would cost marketplace points. Supports two modes: create new skills from scratch with full validation, or grade/audit existing skills with actionable fix suggestions.

Prerequisites

• Claude Code CLI with skill support (v2.1.78+ for advanced features like effort, maxTurns)
• Python 3.10+ for validation scripts (validate-skill.py, aggregate_benchmark.py)
• Target skill directory writable (~/.claude/skills/ or .claude/skills/)

Instructions

Mode Detection

Determine user intent from their prompt:

• Create mode: "create a skill", "build a skill", "new skill" -> proceed to Step 1
• Validate mode: "validate", "check", "grade", "score", "audit" -> jump to Validation Workflow

Communicating with the User

Pay attention to context cues to understand the user's technical level. Skill creator is used by people across a wide range of familiarity — from first-time coders to senior engineers. In the default case:

• "evaluation" and "benchmark" are borderline but OK
• For "JSON" and "assertion", check for cues the user knows these terms before using them without explanation
• Briefly explain terms if in doubt

Step 1: Understand Requirements

If the current conversation already contains a workflow the user wants to capture (e.g., "turn this into a skill"), extract answers from the conversation history first — the tools used, the sequence of steps, corrections the user made, input/output formats observed. Confirm with the user before proceeding.

Ask the user with AskUserQuestion:

Skill Identity:

• Name (kebab-case, gerund preferred: processing-pdfs, analyzing-data)
• Purpose (1-2 sentences: what it does + when to use it)

Execution Model:

• User-invocable via /name? Or background knowledge only?
• Accepts arguments? ($ARGUMENTS substitution)
• Needs isolated context? (context: fork for subagent execution)
• Explicit-only invocation? (disable-model-invocation: true — prevents auto-activation, requires /name)

Required Tools:

• Read, Write, Edit, Glob, Grep, WebFetch, WebSearch, Agent, AskUserQuestion, Skill
• Bash must be scoped: Bash(git:*), Bash(npm:*), etc.
• MCP tools: ServerName:tool_name

Optional: declare disallowed-tools (schema 3.7.0+):

• Defense-in-depth — an explicit denylist layered on top of the allowed-tools allowlist
• Useful when the skill needs broad Bash but should never reach for high-risk operations like rm, curl to arbitrary hosts, .env file edits, or system-config writes
• Example: disallowed-tools: [Bash(rm:*), Bash(curl:*), Bash(wget:*), Bash(sudo:*), Edit(.env), Write(.env)]
• A pattern must never appear in BOTH allowed-tools and disallowed-tools — the marketplace validator reports the overlap as an ERROR
• Default: omit the field (don't clutter frontmatter with a denylist the allowlist already covers)
• Naming parallel: skills use kebab-case disallowed-tools; agents use camelCase disallowedTools — never copy-paste between the two without renaming

Complexity:

• Simple (SKILL.md only)
• With scripts (automation code in scripts/)
• With references (documentation in references/)
• With templates (boilerplate in templates/)
• Full package (all directories)

Location:

• Global: ~/.claude/skills/<skill-name>/
• Project: .claude/skills/<skill-name>/

Step 2: Plan the Skill

Before writing, determine:

Degrees of Freedom:

Think of it as narrow bridge vs open field: a deployment skill is a narrow bridge (one safe path, guard rails everywhere), while a writing skill is an open field (Claude roams freely within broad boundaries). Match constraint level to the task.

Workflow Pattern (see ${CLAUDE_SKILL_DIR}/references/workflows.md):

• Sequential: fixed steps in order
• Conditional: branch based on input
• Wizard: interactive multi-step gathering
• Plan-Validate-Execute: verifiable intermediates
• Feedback Loop: iterate until quality met
• Checklist Workflow: copy-pasteable progress tracking for complex multi-step processes
• Search-Analyze-Report: explore and summarize

Output Pattern (see ${CLAUDE_SKILL_DIR}/references/output-patterns.md):

• Strict template (exact format)
• Flexible template (structure with creative content)
• Examples-driven (input/output pairs)
• Visual (HTML generation)
• Structured data (JSON/YAML)

Step 3: Initialize Structure

Create the skill directory and files:

mkdir -p {location}/{skill-name}
mkdir -p {location}/{skill-name}/scripts      # if needed
mkdir -p {location}/{skill-name}/references   # if needed
mkdir -p {location}/{skill-name}/templates    # if needed
mkdir -p {location}/{skill-name}/assets       # if needed
mkdir -p {location}/{skill-name}/evals        # for eval-driven development

Steps 4-10: Write, Validate, Test, Iterate, Optimize, Report

For detailed guidance on writing SKILL.md (frontmatter rules, description scoring, body guidelines, string substitutions, DCI syntax), creating supporting files, validation, testing, iteration, description optimization, and final reporting, see Creation Guide.

Key rules:

• version, author, license, compatibility, and tags are top-level fields (not nested under metadata:)
• Scope Bash: Bash(git:*) not bare Bash
• Optional disallowed-tools denylist (schema 3.7.0+) must never overlap allowed-tools (validator ERROR)
• Keep under 500 lines; offload to references/ if longer
• Include "Use when" and "Trigger with" in description for enterprise scoring
• No XML tags in name or description (Anthropic spec prohibition)
• No time-sensitive information; use 'old patterns' section for deprecated approaches
• Include feedback loops for quality-critical workflows
• Run python3 ${CLAUDE_SKILL_DIR}/scripts/validate-skill.py --grade {skill-dir}/SKILL.md to validate
• Create evals/evals.json with 3+ scenarios, iterate until all assertions pass