A committed auth.json hands out your private-package tokens, secure-http false lets anyone on the network swap a package, and laravel-debugbar in require ships a debug panel to production. This skill finds those and six other Composer mistakes before your next install, and tells you how to fix each one.

Your agent runs a scanner on composer.json, composer.lock and Composer's auth.json. It knows Composer's own rules: php and ext-* lines are platform requirements, Composer 2 repositories are canonical by default, and the default package type is library. It reads .gitignore so an ignored auth.json is not reported. It never runs Composer or downloads anything. It returns BLOCK, REVIEW, or NOT FLAGGED, with the file, line, reason, and a safer approach. Tokens are never printed.

What it checks

BLOCK (fix before shipping)

• K01: an auth.json with tokens or passwords that git does not ignore

• K02: secure-http false, disable-tls, or http:// and git:// package sources

• K03: tokens or passwords written in composer.json

REVIEW (check before release)

• K11: Composer scripts that download and run code

• K12: allow-plugins set to true

• K13: *, open-ended or dev-branch versions, or minimum-stability dev

• K14: a project without composer.lock

• K15: repositories with canonical false (dependency confusion)

• K16: debug and test tools in require instead of require-dev

What's included

• The scanner (Python 3.9+, standard library only)

• A rules guide with the reason and safer approach for each rule

• A risky and a safe example project, and the real report the scanner produces on the risky one

Limits

It reads each file as written: known vulnerabilities in package versions (use composer audit), the global Composer config, COMPOSER_AUTH and what plugins do are not checked. NOT FLAGGED is not a guarantee.

Contents

One ZIP file with 11 files: SKILL.md, the scanner in scripts/ (3 Python files), references/rules.md (every rule with its reason and safer approach), and examples/ (5 example files plus report.md, the real scanner output on the risky example).

How to use: This is an agent skill: a folder with a SKILL.md file and Python scripts that an AI coding agent such as Claude Code, Cursor, or Codex CLI can run. Unzip it into your agent's skills folder (for Claude Code: ~/.claude/skills/), then ask your agent to check your project. The scripts also run on their own with Python 3.9 or newer, with nothing to install. They read your files only: no network access, and nothing is sent anywhere.