agentleFS
Sign inStart free
mykaelalexandremykaelalexandre/agent-skills/skills/owasp-security-review/SKILL.md

owasp-security-review

Skill0 starsChanged 23 days agoMITView on GitHub

Review code or a diff for security weaknesses using the OWASP Top 10 (2021) and ASVS chapters as a checklist, and report evidenced findings with severity and fixes. Use when the user asks for a security review, audit, threat check or "is this safe" on application code, an API or a pull request.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. jgamaraalv/delivery-loop★ 0same name
    2. Smaui94/ClaudeSkillCollection★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Only review code you can read.

    line 16
  2. Distinctive:

    Every finding needs a file path and line, a short evidence quote, and an explanation of how an attacker reaches it.

    line 38
  3. Very specific:

    Do not invent CVE ids or version vulnerabilities.

    line 40
  4. Distinctive:

    Never paste real secrets into the report.

    line 42
  5. Distinctive:

    Prefer few, real findings over many speculative ones, but list speculative ones under "Needs verification" when the risk would be high.

    line 43
  6. Distinctive:

    Do not modify code unless the user asks for fixes.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

OWASP security review

A structured, evidence-first review. Output is a list of findings a developer can act on, not a generic lecture. Reference checklists are in references/, one file per category.

Scope first

  1. Confirm what is in scope: a directory, a diff, one feature. Ask if unclear. Note the stack (language, framework, datastore, auth mechanism) and where untrusted input enters (routes, queue consumers, file uploads, CLI args, webhooks).
  2. Only review code you can read. Do not run exploits against live systems, and do not run the application unless the user asks and the environment is isolated.

Procedure

  1. Map the surface. List entry points, auth boundaries, data stores, outbound calls, secrets handling and dependency manifests. Keep it short.
  2. Walk the categories in order, A01 to A10. For each, open the matching file in references/ and apply its "What to look for" list to the in-scope code. Use search (grep) for the sink patterns listed there, then read the surrounding code to see whether input really reaches them.
  3. Confirm before reporting. A pattern match is a lead, not a finding. Trace the data flow from source to sink, check for validation or framework protection upstream, and decide: confirmed, likely (needs runtime or config you cannot see) or discarded.
  4. Rate severity using impact times exploitability for this app: Critical, High, Medium, Low, Info. Say why in one clause.
  5. Propose the fix as a concrete change (a code snippet or config line), using the safe pattern from the reference file adapted to the project's stack.
  6. Report in the format below. Finish with categories that had no findings, and the things you could not assess (deployment config, secrets in CI, infrastructure).

Rules

  • Every finding needs a file path and line, a short evidence quote, and an explanation of how an attacker reaches it. No finding without evidence.
  • Do not invent CVE ids or version vulnerabilities. For dependencies, report age and missing lockfiles from what you can read, and recommend running an audit tool (pip-audit, npm audit).
  • Never paste real secrets into the report; show the first and last two characters only.
  • Prefer few, real findings over many speculative ones, but list speculative ones under "Needs verification" when the risk would be high.
  • Do not modify code unless the user asks for fixes. If asked, fix one finding at a time and keep tests passing.
  • This is a review aid, not a certification. Say that it does not replace testing or a human security review for high-risk systems.

Report format

# Security review: <scope>
Stack: ... · Reviewed: <files/dirs> · Method: manual code review against OWASP Top 10 (2021)

## Summary
<2-3 sentences; counts by severity>

## Findings
### [SEVERITY] A03 SQL injection in /search  (confirmed)
- Where: app.py:59
- Evidence: `db.execute(f"... LIKE '%{q}%'")`
- Impact: attacker-controlled SQL; can read or modify any table.
- Fix: bind the parameter: `db.execute("... LIKE ?", (f"%{q}%",))`

## Needs verification
## Categories with no findings
## Not assessed

Category index

IdCategoryReference
A01Broken Access Controlreferences/A01-broken-access-control.md
A02Cryptographic Failuresreferences/A02-cryptographic-failures.md
A03Injectionreferences/A03-injection.md
A04Insecure Designreferences/A04-insecure-design.md
A05Security Misconfigurationreferences/A05-security-misconfiguration.md
A06Vulnerable and Outdated Componentsreferences/A06-vulnerable-outdated-components.md
A07Identification and Authentication Failuresreferences/A07-identification-authentication-failures.md
A08Software and Data Integrity Failuresreferences/A08-software-data-integrity-failures.md
A09Security Logging and Monitoring Failuresreferences/A09-logging-monitoring-failures.md
A10Server-Side Request Forgeryreferences/A10-server-side-request-forgery.md

Sources: https://owasp.org/Top10/ and https://owasp.org/www-project-application-security-verification-standard/. The reference files are my own summaries, not OWASP text; read the originals for authoritative detail.

A worked target is in examples/vulnerable-app/ (never deploy it) with the answer key in examples/expected-findings.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free