One "SELECT ... WHERE id = $id" or one echo $_GET['q'] is enough for someone to read your database or take over your users' sessions. This skill finds those and twelve other PHP security mistakes before you deploy, and tells you how to fix each one.

Your agent runs a scanner on your .php files. It reads only the code inside PHP tags, skips comments, understands quotes and heredocs, follows a query or command built in a variable (also with .=) to the call that runs it, and follows request values ($_GET, $_POST, $_COOKIE, uploaded file names, $request->input()) through variables. It knows the safe patterns: placeholders and $wpdb->prepare(), (int) casts, escapeshellarg(), htmlspecialchars() and esc_html(), basename(). It never runs PHP. It returns BLOCK, REVIEW, or NOT FLAGGED, with the file, line, reason, and a safer approach. Secrets are never printed.

What it checks

BLOCK (fix before shipping)

• P01: SQL built from variables and run with mysqli_query, PDO, $wpdb or raw Laravel queries

• P02: exec, system, shell_exec or backticks with unescaped values

• P03: eval, create_function or preg_replace /e

• P04: include or require of a path from the request

• P05: unserialize on request or cookie data

• P06: a hard-coded password, API key or token (variables, define, config arrays, env() defaults, connect calls)

• P07: request values printed with echo, print or <?= without escaping (XSS)

REVIEW (check before release)

• P11: md5, sha1 or crypt on passwords

• P12: TLS certificate checks turned off (cURL, Guzzle, stream contexts)

• P13: display_errors on or phpinfo()

• P14: redirects to a URL from the request

• P15: rand, mt_rand or uniqid for tokens, OTPs and reset codes

• P16: extract($_POST) or parse_str with one argument

• P17: file paths from the request or an uploaded file's own name

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 code as text, not as a full PHP parser: values are followed within one file and in file order, not across functions or includes. Templates (Blade, Twig) and vendor code are not checked. NOT FLAGGED is not a guarantee.

Contents

One ZIP file with 16 files: SKILL.md, the scanner in scripts/ (3 Python files), references/rules.md (every rule with its reason and safer approach), and examples/ (10 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.