peruse

peruse is a skill for Claude Code from iroha924/mumei. It costs 160 tokens per session (7,141 once invoked), scanned A, original, MIT.

A read-only review pipeline for a planned software feature. It runs security and adversarial checks, validates reported problems, combines the results, and saves a review record.

In plain words
What is it for?
Use it after all tasks in a plan are complete and a review is pending. It is intended for features managed in a Mumei plan, a project workflow for tracking planned changes.
Why use it?
It finds security and implementation concerns before a planned feature is accepted. It also filters out findings that cannot be confirmed.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions subagents; mentions Codex.

Runs only inside its plugin — its command needs a path that Claude Code sets for a plugin’s own hooks and for nothing else. Install the plugin, not this.

Part of the mumei plugin — 11 skills, 10 agents, 18 hooks shipped together

Good fit Use it after all tasks in a plan are complete and a review is pending. It is intended for features managed in a Mumei plan, a project workflow for tracking planned changes.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add iroha924/mumei
Claude Code
/plugin install mumei

Made for: Claude Code.

Or install mumei, the plugin that ships this one along with the rest of its 11 skills, 10 agents, 18 hooks.

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 peruse

README.md
[![agentmods](https://agentmods.dev/badge/skills/iroha924/mumei/peruse/github.svg)](https://agentmods.dev/skills/iroha924/mumei/peruse)
Your own site
<a href="https://agentmods.dev/skills/iroha924/mumei/peruse"><img src="https://agentmods.dev/badge/skills/iroha924/mumei/peruse/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 peruse

Your own site · 80×15
<a href="https://agentmods.dev/skills/iroha924/mumei/peruse"><img src="https://agentmods.dev/badge/skills/iroha924/mumei/peruse.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 160 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,141 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 $0.00160 $0.07141
Opus 5 $0.00080 $0.03571
Sonnet 5 $0.00032 $0.01428
Haiku 4.5 $0.00016 $0.00714

Measured 9d ago against content hash 6074e2df32db, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

peruse 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 9d 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.

skills/peruse/SKILL.md · 546 lines

How it starts

The opening of the file, as written. The whole thing — 546 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Peruse — plan-vehicle review pipeline

This skill is the plan-vehicle counterpart of Phase 5 in /mumei:compose. It runs only against plan-vehicle features (state.json under .mumei/plans/<slug>/). For spec-vehicle review, use /mumei:compose (which drives the same pipeline as part of its lifecycle).

When to use

  • After completing all TaskCreate/TaskCompleted work in a Claude plan-mode session, when pending_review=true is set in .mumei/plans/<slug>/state.json.
  • The Stop hook (L-R1) and the git push PreBash hook (L-R2) will block session-end / push until a passing review JSON exists.

When NOT to use

  • For spec-vehicle features (run /mumei:compose instead).
  • Before all planned tasks have been marked completed (the skill aborts with a hint message).
  • For projects without an active mumei feature.

Method

All steps below assume the current working directory is the project root and a git repo. The skill is conservative: it never edits source files, never spawns commits, and never auto-archives.

Step 1 — Resolve active feature and refuse non-plan-vehicle invocations

source "${CLAUDE_PLUGIN_ROOT}/hooks/_lib/state.sh"
source "${CLAUDE_PLUGIN_ROOT}/hooks/_lib/review.sh"
source "${CLAUDE_PLUGIN_ROOT}/hooks/_lib/ledger.sh"
source "${CLAUDE_PLUGIN_ROOT}/hooks/_lib/residual.sh"
source "${CLAUDE_PLUGIN_ROOT}/hooks/_lib/cost-log.sh"

slug="$(mumei_current_feature 2>/dev/null || true)"
if [[ -z "$slug" ]]; then
  echo "no active mumei feature (.mumei/current is empty); aborting" >&2
  exit 0
fi

if ! mumei_state_is_plan_vehicle "$slug"; then
  echo "/mumei:peruse is plan-vehicle only. Active feature '${slug}' is a spec-vehicle feature; use /mumei:compose to drive its review (Phase 5)." >&2
  exit 0
fi

review_dir=".mumei/plans/${slug}/reviews"
state_path=".mumei/plans/${slug}/state.json"

Read the full file on GitHub · 546 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. 9d ago First seen · 546 lines · 160 tokens per session scan A 6074e2df32db

Subscribe to this mod's changes

peruse is a skill published in the GitHub repository iroha924/mumei (2 stars, last pushed 3d ago), licensed MIT. It adds 160 tokens to every session and 7,141 once invoked, about $0.0008 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.

Related

Other skills, from other repositories

evaluator-review-contract

Internal Auto-Harness evaluator skill for sprint contract review before implementation. Use only inside the Evaluator subagent during review mode.

redker56/auto-harness · 30 tokens

md-audit

Read-only code quality audit — scan the current working directory for common issues (bugs, dead code, security hotspots, missing error handling) and return a prioritised findings report. No files are edited. Use when asked to "audit the code", "quick audit", "find issues", "code scan", or "what's wrong with this…

chaitanyagiri/munder-difflin · 85 tokens

harness

A meta-skill for designing and maintaining a project harness: a coordinated set of specialist agents, skills, and project instructions. It defines their roles, connects them to the project, and updates the setup as the work changes.

revfactory/harness · 170 tokens

load-pr-review

Load GitHub PR review comments into AI session — analyze, triage, plan. Default: analysis-only (no auto-fix). Use when: reviewing PR feedback, planning fixes, addressing review comments, replying to reviewers. Not for: creating reviews (use codex-review-fast), creating PRs (use create-pr), viewing PR status (use…

sd0xdev/sd0x-harness · 77 tokens

issue-analyze

GitHub Issue and PR review thread deep analysis with Codex blind verdict. Use when: analyzing issue root cause, classifying problems, investigation planning, triaging PR review comments for actionability. Not for: fixing bugs (use bug-fix), code exploration (use code-explore). Output: classified analysis + verdict…

sd0xdev/sd0x-harness · 73 tokens

best-practices

Industry best practices conformance audit with mandatory adversarial debate. Produces audit artifact: verdict (OK/WARN/FAIL) + gap roadmap + debate proof. Use when: auditing current implementation against industry standards, checking compliance with best practices, benchmarking implementation quality, verifying a…

sd0xdev/sd0x-harness · 101 tokens