skill-fleet-review

skill-fleet-review is a skill for Claude Code from YoungjaeDev/my-claude-plugins. It costs 178 tokens per session (2,531 once invoked), scanned A, original, MIT.

A review tool that measures every coding-agent skill in a plugin tree, selects the skills needing attention, and produces an audit report and CSV.

In plain words
What is it for?
Running a fleet-wide review before or after reorganising plugins, choosing refactoring priorities, and tracking P0, P1, and P2 findings.
Why use it?
It finds repository-wide problems without requiring a reviewer to read every skill in full, while still checking descriptions for overlapping triggers.

Skill for Claude Code

Written for Claude Code: ${CLAUDE_PLUGIN_ROOT variable. Also seen: reads .claude/ paths; mentions subagents; positional $N argument.

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 docs plugin — 8 skills shipped together

Good fit Running a fleet-wide review before or after reorganising plugins, choosing refactoring priorities, and tracking P0, P1, and P2 findings.

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 YoungjaeDev/my-claude-plugins
Claude Code
/plugin install docs

Made for: Claude Code.

Or install docs, the plugin that ships this one along with the rest of its 8 skills.

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-fleet-review

README.md
[![agentmods](https://agentmods.dev/badge/skills/youngjaedev/my-claude-plugins/skill-fleet-review/github.svg)](https://agentmods.dev/skills/youngjaedev/my-claude-plugins/skill-fleet-review)
Your own site
<a href="https://agentmods.dev/skills/youngjaedev/my-claude-plugins/skill-fleet-review"><img src="https://agentmods.dev/badge/skills/youngjaedev/my-claude-plugins/skill-fleet-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-fleet-review

Your own site · 80×15
<a href="https://agentmods.dev/skills/youngjaedev/my-claude-plugins/skill-fleet-review"><img src="https://agentmods.dev/badge/skills/youngjaedev/my-claude-plugins/skill-fleet-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 178 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,531 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 $0.00178 $0.02531
Opus 5 $0.00089 $0.01265
Sonnet 5 $0.00036 $0.00506
Haiku 4.5 $0.00018 $0.00253

Measured 5d ago against content hash 1d2e48fec1e7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

skill-fleet-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 5d 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/docs/skills/skill-fleet-review/SKILL.md · 181 lines

How it starts

The opening of the file, as written. The whole thing — 181 lines — stays where its author put it; the contents beside it link to each section on GitHub.

skill-fleet-review

Review every skill in the tree without reading every skill with a model.

Overview

A fleet review that reads 50 bodies with a model costs more than it returns, and most of what it would find is mechanical. So: measure everything, let the numbers pick the cohort, and spend judgment there.

The pointer axis is the exception. Descriptions are short, so reading all of them is cheap, and a trigger collision is invisible from inside one skill. It only shows up when siblings are read together, which is exactly what a fleet pass makes possible.

The rules being applied belong to docs:skill-forge and live in its bundled references.

When to use

A repository-wide sweep: before or after reorganizing plugins, when choosing what to refactor next, or on a periodic pass over accumulated drift.

Do not use it for one skill (docs:skill-audit) or to write one (docs:skill-forge).

Prerequisites

Run from the repository root. Node 18+ for the measurement script. Nothing else: no MCP server, no external CLI, and no other skill is required at any step.

Procedure

1. Measure the fleet

# Honor a caller-supplied PLUGIN_ROOT first — the abort message below tells the user to
# export it, and starting from CLAUDE_PLUGIN_ROOT would overwrite that escape hatch.
PLUGIN_ROOT="${PLUGIN_ROOT:-${CLAUDE_PLUGIN_ROOT:-}}"
[ -z "$PLUGIN_ROOT" ] && [ -d plugins/docs/skills ] && PLUGIN_ROOT=plugins/docs
if [ -z "$PLUGIN_ROOT" ]; then
  # Rank Codex cache candidates on the version basename, not the whole path: a plain
  # sort puts 0.9.0 above 0.10.0 and lets the marketplace directory outrank the version.
  cache_root="${CODEX_PLUGIN_CACHE:-$HOME/.codex/plugins/cache}"
  if sort -V </dev/null >/dev/null 2>&1; then vsort="sort -V"; else vsort="sort -t. -k1,1n -k2,2n -k3,3n"; fi
  PLUGIN_ROOT=$(ls -1d "$cache_root"/*/docs/* 2>/dev/null | awk -F/ '{print $NF "\t" $0}' | $vsort | tail -1 | cut -f2-)
fi
[ -d "$PLUGIN_ROOT/skills" ] || { echo "docs plugin root not found; export PLUGIN_ROOT" >&2; exit 1; }
MEASURE="$PLUGIN_ROOT/skills/skill-forge/scripts/measure-skills.mjs"
mkdir -p docs/audit
# Name the roots to sweep. With no argument the measurer defaults to `plugins/` when it
# exists, so a repository that also keeps skills under `.claude/skills/` gets a partial
# scan reported as repository-wide. List every root that holds skills here.
SWEEP_ROOTS="${SWEEP_ROOTS:?set SWEEP_ROOTS to every skill root to sweep, e.g. \"plugins .claude/skills\"}"
# zsh does not word-split an unquoted parameter, so a multi-root value would arrive as one
# path and measure nothing. Ask for POSIX splitting; the probe is a no-op under bash/sh.
set -o shwordsplit 2>/dev/null || true
node "$MEASURE" $SWEEP_ROOTS; rc=$?
# Exit 2 means part of the tree could not be read. Continuing would build a cohort and a
# report from a baseline that silently omits those skills — the false-clean sweep this
# skill's Pitfalls section forbids. Stop and report the unverified scope instead.
[ "$rc" -eq 0 ] || { echo "measure-skills exited $rc — scope incomplete, sweep stopped" >&2; exit 1; }

# Never overwrite an earlier run's artifacts: the first run's numbers are the baseline
# the next one is compared against. Testing for the file and then redirecting is not
# enough — two sweeps started in the same second, or in parallel, both pass the test and
# the second `>` wins. Reserve the CSV atomically instead: under `noclobber` the shell
# opens with O_EXCL, so exactly one racing run gets each candidate.
# Both artifacts share one base, so the base is what gets reserved — not just the CSV.
# Checking only `.csv` re-picked a base whose `.md` still existed (the failure cleanup
# below removes the CSV but never the report), and Step 6 then overwrote that report.
base="docs/audit/$(date +%Y-%m-%d)-fleet"; OUT="$base"; n=0
until [ ! -e "$OUT.md" ] && (set -o noclobber; : > "$OUT.csv") 2>/dev/null; do
  n=$((n + 1)); OUT="$base-$n"
  [ "$n" -gt 50 ] && { echo "measure: cannot reserve an audit path under $base" >&2; exit 1; }
done
node "$MEASURE" --csv $SWEEP_ROOTS > "$OUT.csv"; rc=$?
# Remove the reservation on failure. Leaving the truncated file behind hands the next
# reader a zero-row CSV that looks exactly like a fleet with nothing to report.
[ "$rc" -eq 0 ] || { rm -f "$OUT.csv"; echo "measure-skills --csv exited $rc — reservation $OUT.csv removed, no baseline written" >&2; exit 1; }
echo "report base: $OUT"

Read the full file on GitHub · 181 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. 5d ago Changed · -4 tokens per session 1d2e48fec1e7
  2. 10d ago First seen · 181 lines · 182 tokens per session scan A fba0dcf364f4

Subscribe to this mod's changes

skill-fleet-review is a skill published in the GitHub repository YoungjaeDev/my-claude-plugins (2 stars, last pushed 3d ago), licensed MIT. It adds 178 tokens to every session and 2,531 once invoked, about $0.0009 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

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

insight-error-page

Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…

vercel/next.js · 83 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens