Security Guidance
Security scan passedAnthropic's security review for Claude Code: flags injection, XSS, secrets and more as Claude edits, finishes a turn and commits.

Alireza Rezvani's Claude Code hook that stops Claude before it writes common security anti-patterns, such as eval, pickle or SQL f-strings.
by alirezarezvani· SecurityUnclaimed · Yours? Claim it
4 images
About
Security Guidance (claude-skills) is a community plugin from Alireza Rezvani's claude-skills collection. It's a hook for Claude Code: before Claude edits or writes a file, it checks the change against a list of known-dangerous patterns, and when one matches it shows Claude a warning and blocks that edit. It's ported from David Dworken's MIT-licensed hook, with three patterns added. It isn't Anthropic's own Security Guidance plugin. What it catches - Command injection: child_process.exec, exec( and execSync(, os.system, and subprocess with shell=True - Code injection: eval( and new Function - XSS: dangerouslySetInnerHTML, document.write and .innerHTML = - Unsafe deserialization: pickle, yaml.load( and yaml.unsafe_load - SQL injection: SELECT, INSERT, UPDATE or DELETE queries built with f-strings or .format( - GitHub Actions workflow files, where untrusted ${{ }} inputs can inject commands How it works - It runs before every Edit, Write or MultiEdit, with no setup. It needs only Python 3, with no other dependencies. - A match blocks the edit and tells Claude why. The same warning for the same file isn't repeated in that session, so Claude can go ahead once it has seen it. - For a deliberate use (say, pickle in an internal tool), the author suggests a comment in the file explaining why it's safe. How to install it In Claude Code, run /plugin marketplace add alirezarezvani/claude-skills, then /plugin install security-guidance@claude-code-skills. Good to know - It's a community plugin, not Anthropic's. - The checks are plain text matches, so expect some false positives. For stricter checks, the author suggests adding a SAST tool such as Semgrep or CodeQL to CI. - Turn it off for one session with ENABLE_SECURITY_REMINDER=0. - It keeps per-session state and a debug log in ~/.claude/.
Once it’s added
“Add a security hook that blocks unsafe code before it's written.”
“Detect command injection before you write the file.”
“Prevent SQL injection patterns in this project.”
“Warn me whenever code uses eval, pickle or os.system.”
“Add a GitHub Actions security hook.”
Before you add it
Reads only
The file path and content of each Edit, Write or MultiEdit, before it runs
Makes changes · your agent can ask first
Blocks a matching edit once per file and rule, with a warning for Claude · Keeps session state and a debug log in ~/.claude/
From people who use it
Step by step
Plugins people install alongside Security Guidance (claude-skills).
No reviews yet. Used Security Guidance (claude-skills)? Say how it went.
Write a reviewRuns on your machine.
Runs on your machine.
/plugin marketplace add alirezarezvani/claude-skills /plugin install security-guidance@claude-code-skills
Scanned and passed · checked 6 h ago
Nothing new reaches your agent until a person has looked at it.
Scanned on every change, and again when our rules improve.
Tell us and a person on the trust team will look at it. Or email [email protected].
Nobody has shared how they use Security Guidance (claude-skills) yet.
Security scan
Passed
Safe to install in every agent it supports. Claude plugins.
Plugins people install alongside Security Guidance (claude-skills).
Anthropic's security review for Claude Code: flags injection, XSS, secrets and more as Claude edits, finishes a turn and commits.
Equips your coding agent with SAST, dependency scanning, a phased hardening workflow and STRIDE threat-modeling skills.
Microsoft's skills for Azure Data Explorer: build Kusto graphs, compose IRQL threat hunts and turn results into visual graphs.
CrowdStrike's local MCP server for Falcon: your agent triages detections, looks up hosts and threat intel, and runs Next-Gen SIEM searches.