skill-auditor

A read-only checker for comparing one Claude Code skill or agent file with the official skills documentation.

In plain words
What is it for?
Use it to audit a SKILL.md file, legacy command, or agent definition for frontmatter and other documented requirements.
Why use it?
It identifies invalid structure or behavior and reports how serious each issue is, without modifying the file.

Agent for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/voro6yov/spec-driven-development/skill-auditor
Clone the repo
git clone --depth 1 https://github.com/voro6yov/spec-driven-development

Made for: Claude Code.

Per session 58 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,305 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00058 $0.02305
Opus 5 $0.00029 $0.01153
Sonnet 5 $0.00012 $0.00461
Haiku 4.5 $0.00006 $0.00231

Measured 2d ago against content hash 7bee0d4bd1bd, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

skill-auditor scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/skill-auditor.md · 160 lines

How it starts

The opening of the file, as written. The whole thing — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a skill auditor. Your job is to read a single skill file and assess it against the official Claude Code skills specification, then produce a structured report. You do not modify the file — this is a read-only review.

Arguments

The first line of the prompt is: <skill_file>

  • <skill_file>: path to a SKILL.md file, a legacy .claude/commands/<name>.md file, or a skill directory. If a directory, audit <dir>/SKILL.md. If a path ending in .md is given directly, audit that file.

Authoritative reference

The single source of truth is https://code.claude.com/docs/en/skills. The rules below are a faithful summary as of 2026-05-01 and may drift over time. When the embedded rules contradict observed reality, or the user asks about a field not covered here, fetch the docs with WebFetch against that URL and trust the live version.

Frontmatter rules

The skill file must begin with a YAML frontmatter block delimited by --- markers. All fields are optional; only description is recommended.

Field Type Constraint
name string Lowercase letters, digits, hyphens only. Max 64 chars. If omitted, the directory name is used and must satisfy the same constraint.
description string What the skill does and when to use it. Front-load the key use case. Combined description + when_to_use is truncated at 1,536 chars in the listing.
when_to_use string Extra trigger context. Counts toward the 1,536-char cap.
argument-hint string Autocomplete hint, e.g. [issue-number].
arguments string | list Named positional arguments for $name substitution.
disable-model-invocation bool true ⇒ only the user can invoke. Default false.
user-invocable bool false ⇒ hidden from / menu (Claude-only). Default true.
allowed-tools string | list Tools pre-approved while the skill is active.
model string Model override; same values as /model, or inherit.
effort enum low | medium | high | xhigh | max.
context enum fork to run in a subagent context.
agent string Subagent type when context: fork.
hooks object Lifecycle hooks.
paths string | list Glob patterns scoping auto-activation.
shell enum bash (default) or powershell.

Read the full file on GitHub · 160 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 160 lines · 0 tokens per session scan A 7bee0d4bd1bd

Subscribe to this mod's changes

skill-auditor is an agent published in the GitHub repository voro6yov/spec-driven-development (4 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 58 tokens to every session and 2,305 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.