security-review skillA
security-review is agent-read markdown (skill) from jamalmohafil/claude-skills: Run a security review of code changes exactly like Claude Code's /security-review command — but in ANY AI coding agent (Claude Code, Cursor, Codex, Windsurf, Gemini CLI, Cline…). Reviews the pending branch diff (or a specific PR, uncommitted changes, or a whole file/folder) for HIGH-CONFIDENCE, actually-exploitable vulnerabilities — SQL/command/template/NoSQL injection, path traversal, auth & authorization bypass, privilege escalation, hardcoded secrets, weak crypto, insecure deserialization / R.
Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.
What the file says
# security-review — Claude Code's `/security-review`, in any agent You are a **senior security engineer** doing a focused security review of code changes. The whole point of this review is **signal, not noise**: report only vulnerabilities you are **> 80% confident are real and exploitable**. A short report with 2 true bugs beats a long report with 20 maybes — the second one gets ignored and trains people to distrust the tool. This skill mirrors the methodology of Anthropic's official `/security-review` command. Follow the steps below in order. Read the three files in [`reference/`](reference/) — they are not optional; they hold the full taxonomy, the exact filtering rules, and the report format. --- ## The golden rules (read once, apply always) 1. **High confidence only.** Only flag issues where you can point to a concrete, exploitable attack path. If you cannot write a realistic exploit scenario, do not report it. 2. **Only what changed.** Review the **security implications newly introduced** by these changes. Do not comment on pre-existing issues in untouched code. 3. **Impact first.** Prioritize things that lead to unauthorized access, data breach, RCE, or …
Read the whole file at its exact version.
How to install
mdr add jamalmohafil/claude-skills/security-review@git:20260802.08449e2mdr add jamalmohafil/claude-skills/security-review@sha256:9386bb4610a1ac8aPin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.
[](https://markdownregistry.com/a/art_b5mjb6oei7gp2evf)
1 badge views in 30 days
Versions
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (9880 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
jamalmohafil/claude-skills · 26 stars · license MIT · pushed 2026-09-21 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_b5mjb6oei7gp2evf GET https://markdownregistry.com/api/v1/resolve?ref=jamalmohafil/claude-skills/security-review GET https://markdownregistry.com/api/v1/blob/9386bb4610a1ac8a3f343694904ced04fb28d5d8f860caea633c1a5dac2598ca
Your agent does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
More from jamalmohafil/claude-skills
Every file in jamalmohafil/claude-skills