skill-review

skill-review is a skill for Claude Code from jhoblitt/skill-review-claude. It costs 190 tokens per session (1,542 once invoked), scanned A, original, Apache-2.0.

A review skill for the written instructions used by coding agents, including skills, agents, plugins, and marketplace files. It checks four areas: duplicated rules, task ordering, token use, and security.

In plain words
What is it for?
It is for reviewing instruction changes before committing, pushing, or opening a pull request, or for auditing a repository on request.
Why use it?
It helps catch instruction problems that ordinary code review may miss, such as conflicting copies of a rule, unnecessary serial work, wasteful instructions, or unsafe trust in supplied content.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: mentions CLAUDE.md; mentions subagents; mentions Claude Code.

Part of the skill-review plugin — 1 skill, 4 agents shipped together

Good fit It is for reviewing instruction changes before committing, pushing, or opening a pull request, or for auditing a repository on request.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/jhoblitt/skill-review-claude/skill-review
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.

Any agent
npx skills add jhoblitt/skill-review-claude --skill skill-review
Clone the repo
git clone --depth 1 https://github.com/jhoblitt/skill-review-claude

Made for: Claude Code.

Or install skill-review, the plugin that ships this one along with the rest of its 1 skill, 4 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for skill-review

README.md
[![agentmods](https://agentmods.dev/badge/skills/jhoblitt/skill-review-claude/skill-review/github.svg)](https://agentmods.dev/skills/jhoblitt/skill-review-claude/skill-review)
Your own site
<a href="https://agentmods.dev/skills/jhoblitt/skill-review-claude/skill-review"><img src="https://agentmods.dev/badge/skills/jhoblitt/skill-review-claude/skill-review/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for skill-review

Your own site · 80×15
<a href="https://agentmods.dev/skills/jhoblitt/skill-review-claude/skill-review"><img src="https://agentmods.dev/badge/skills/jhoblitt/skill-review-claude/skill-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 190 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,542 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00190 $0.01542
Opus 5 $0.00095 $0.00771
Sonnet 5 $0.00038 $0.00308
Haiku 4.5 $0.00019 $0.00154

Measured 11d ago against content hash 0d068529ea42, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

skill-review 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 11d 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.

plugins/skill-review/skills/skill-review/SKILL.md · 116 lines

How it starts

The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Skill review — drift, concurrency, token-usage, and security gate

Instruction files fail in four ways code does not. A rule restated in two places reads fine in both until one of them moves. A procedure that walks independent work one item at a time reads fine too — it just taxes every executor that follows it. A procedure that spends a frontier model on what a script could do reads fine forever, because prose carries no price tag. And a procedure that reads content an attacker can author, with capability that work never needed in reach, reads fine right up until the attacker writes it. This skill reviews skill, agent, and canon prose the way a maintainer reviews code, on four axes.

When to gate

Nothing fires the gate for you. Run it unprompted as soon as a branch that touched instruction prose is headed for a commit, a push, or gh pr create — a SKILL.md, an agent, command, or references/ file, a CLAUDE.md or AGENTS.md, or a plugin, marketplace, hook, or MCP manifest. The verdict is due before the pull request opens, not after.

Axes

Load only the axes the request calls for. Paths are relative to this skill's directory.

  • drift — is every rule stated exactly once? → references/drift.md, agent skill-review:drift-review
  • concurrency — does the control flow match the dependencies? → references/concurrency.md, agent skill-review:concurrency-review
  • token usage — is every operation on the cheapest sufficient mechanism? → references/token-usage.md, agent skill-review:token-usage-review
  • security — can anything the procedure reads steer it, or reach beyond it? → references/security.md, agent skill-review:security-review

A pre-PR gate or a repo-wide audit runs all four. A targeted request loads one: "is this rule stated anywhere else?" is drift, "does this parallelize?" is concurrency, "is this wasting tokens?" is token usage, "could a hostile issue body steer this?" is security.

Shared review contract

Read the full file on GitHub · 116 lines

Files

What ships with it

4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 11d ago First seen · 116 lines · 190 tokens per session scan A 0d068529ea42

Subscribe to this mod's changes

skill-review is a skill published in the GitHub repository jhoblitt/skill-review-claude (2 stars, last pushed 27d ago), licensed Apache-2.0. It adds 190 tokens to every session and 1,542 once invoked, about $0.0010 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.