Skill Evaluator

Discover all skills in the project, score them across 6 criteria, and infer the appropriate effort level based on content analysis.

When to Use

• New project: run once to establish baseline quality
• Before committing a skill to a team repo
• After bulk-importing skills from another project
• When adding effort fields for the first time (v2.1.80+)

What Gets Audited

All SKILL.md files and flat .md files found in:

• .claude/skills/**
• ~/.claude/skills/** (if requested)
• Any path passed as argument: /eval-skills ./my-skills-dir

---

Scoring Criteria (14 pts per skill)

#	Criterion	Max	What is checked
1	name	1	Present, lowercase, hyphens only, matches directory name
2	description	2	Present + has "Use when" / "when to" / trigger phrasing
3	allowed-tools	2	Present + not overly broad (Bash without scoping when read-only)
4	effort	3	Present (1pt) + appropriate for content (2pt based on inference)
5	content structure	4	Has Purpose/When section (1), has examples/usage (1), has clear workflow (1), no placeholder text (1)
6	bonus	+2	argument-hint present (1), version/author metadata (1)

Note: tags is NOT an officially supported frontmatter field in Claude Code. It is ignored by the runtime. Do not include it or score it as a quality criterion.

Thresholds:

• ✅ Good: ≥11/14 (≥80%)
• ⚠️ Needs work: 8–10/14 (60–79%)
• ❌ Fix: <8/14 (<60%)

---

Effort Level Inference Engine

For each skill, analyze description + content and classify using these signals:

low — Mechanical execution, no design decisions

Signals:

• Verbs: commit, push, sync, scaffold, generate (template-based), format, rename, bump, wrap, convert
• No reasoning required: sequential steps, template instantiation, data fetching
• allowed-tools: Bash only, or Read-only
• No sub-agents spawned
• Short workflow (<5 steps)

Examples: /commit, /release-notes, /scaffold, /sync, /format

medium — Analysis with bounded scope, categorization

Signals:

• Verbs: review, triage, analyze, categorize, suggest, evaluate (single file or bounded scope)
• Requires pattern recognition but not architectural reasoning
• allowed-tools: Read + Grep + Bash combination
• May spawn 1-2 sub-agents but with predefined scope
• Produces structured output (tables, categorized lists)

Examples: /code-review (single PR), /issue-triage, /dependency-audit, /test-coverage

high — Design decisions, adversarial reasoning, cross-system analysis

Signals:

• Verbs: architect, redesign, threat-model, audit (security), orchestrate (multi-agent), score, assess trade-offs
• Requires reasoning about edge cases, attack vectors, or system-wide implications
• allowed-tools: broad access (Read + Write + Bash + external tools)
• Spawns multiple sub-agents or uses parallel execution
• Produces analysis with explicit uncertainty or trade-off sections
• Keywords in content: "security", "architecture", "adversarial", "pipeline", "threat", "design decision"

Examples: /security-audit, /architecture-review, /cyber-defense, /eval-agents

Mismatch flag

If a skill has effort: already set but the inferred level differs, flag it:

⚠️ Effort mismatch: declared low, inferred high — skill spawns 4 sub-agents and performs security analysis

---

Execution Instructions

Step 1 — Discovery

# Find all SKILL.md files
find .claude/skills -name "SKILL.md" 2>/dev/null

# Find flat skill files
find .claude/skills -maxdepth 1 -name "*.md" ! -name "README*" 2>/dev/null

# If argument provided, use that path instead

Step 2 — Parse each skill

For each skill file found:

1. Read the full file
2. Extract YAML frontmatter (between first --- and second ---)
3. Parse: name, description, allowed-tools, effort, argument-hint, version
4. Note presence/absence of each field
5. Read the body content for structure analysis

Step 3 — Score and infer

Apply the scoring criteria above to each skill:

• Check frontmatter fields
• Evaluate description quality (does it answer "when to use"? is it under 1024 chars?)
• Evaluate allowed-tools scope (is Bash used when only Read would suffice? are tools scoped with wildcards when possible?)
• Infer effort level from content analysis
• Compare inferred vs declared effort (if set)
• Evaluate content structure (scan for "When to Use", "Purpose", "Example", "Workflow" sections)

Step 4 — Output

Produce a structured report:

# Skills Audit — [project name or path]
Date: [today] | Scanned: N skills

## Summary
| Status | Count |
|--------|-------|
| ✅ Good (≥80%) | N |
| ⚠️ Needs work (60–79%) | N |
| ❌ Fix (<60%) | N |

**Effort coverage**: N/N skills have effort field set

---

## Per-Skill Results

### [skill-name] — [score]/15 [✅/⚠️/❌]

| Criterion | Score | Notes |
|-----------|-------|-------|
| name | ✅ 1/1 | — |
| description | ⚠️ 1/2 | Missing "Use when" phrasing |
| allowed-tools | ✅ 2/2 | Well-scoped |
| effort | ❌ 0/3 | Missing — Recommended: high |
| content structure | ⚠️ 2/4 | No examples section |

**Effort inference**: `high` — skill performs security analysis with adversarial reasoning
  Signals: "threat", "attack surface", "vulnerability scoring" in content; spawns 4 agents

**Priority fixes** (ordered by impact):
1. Add `effort: high` to frontmatter
2. Add "Use when" to description
3. Add a concrete usage example section

---

After all skills: print a Fix Summary — all missing effort fields with recommended values, ready to copy-paste.

---

Fix Summary Format

At the end, print a ready-to-use patch block for all missing/mismatched effort fields:

## Recommended effort fields (copy-paste ready)

skill-name-1: effort: low     # mechanical scaffold
skill-name-2: effort: high    # security analysis, spawns agents
skill-name-3: effort: medium  # code review, bounded scope

And a 1-line count: N skills need effort field · N mismatches · N missing allowed-tools