agentleFS
Sign inStart free
EskyeeEskyee/agentbot-opensource/skills/aeon-vuln-scanner/SKILL.md

aeon-vuln-scanner

Skill17 starsChanged 5 months agoMITView on GitHub

Audit trending repos for real exploitable vulnerabilities and disclose responsibly — Private Vulnerability Reporting for code flaws and verified secrets, public PRs only for already-disclosed dependency CVEs. Semgrep + TruffleHog + osv-scanner + Slither with reachability triage. Skips targets that have no safe disclosure channel. Triggers: "vuln scan owner/repo", "audit this repo", "responsible-disclosure scan", "check for secret leaks", "scan dependencies for CVEs".

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write one sentence: *what attacker controls, what they achieve*.

    line 55
  2. Distinctive:

    Check the call path — reachable from external input in production code?

    line 56
  3. Distinctive:

    Do no harm. No safe channel → no publication.

    line 83
  4. Distinctive:

    Read the code. A scanner hit alone isn't a vulnerability.

    line 84

≈ “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

aeon-vuln-scanner

A scanner that dumps zero-days into public PRs isn't a helper — it's a publisher. This skill triages every finding by reading the code and routes to the right disclosure channel.

Inputs

ParamDescription
varOptional owner/repo. If empty, auto-picks from chained github-trending output or fresh trending API.

Target selection

  • Language: JS/TS, Python, Go, Rust, or Solidity.
  • ≥ 50 stars, not a fork, active in last 6 months.
  • Handles untrusted input (auth, crypto, network, file I/O, templating).
  • Skip: intentionally vulnerable teaching repos (juice-shop, webgoat, *-ctf).
  • Skip if no PVR enabled AND no SECURITY.md — no safe channel.
  • Skip if scanned in last 30 days (dedup via vuln-scanned.json).

Scanners

# Static analysis
semgrep --config=p/security-audit --config=p/owasp-top-ten --config=p/secrets \
  --severity=ERROR --severity=WARNING --json --timeout=300 \
  --exclude=test --exclude=examples --exclude=node_modules .

# Verified secrets (filesystem + git history)
trufflehog filesystem . --only-verified --json
trufflehog git file://. --only-verified --json

# Dependency CVEs across npm/pip/go/cargo/etc
osv-scanner --format=json --recursive .

# Solidity (if .sol files present)
slither . --json out.json --exclude-informational --exclude-low

All-scanners-failed reports error, never clean.

Triage (per candidate)

  1. Open the file at the reported line. Read 30-50 lines of context.
  2. Write one sentence: what attacker controls, what they achieve. Can't? Discard.
  3. Check the call path — reachable from external input in production code?
  4. Drop if in tests, fixtures, examples, behind a feature flag, or requires attacker privs ≥ what's gained.

Disclosure routing

FindingChannel
Dependency CVEPublic PR bumping the dep — CVE already public.
Code vulnerabilityPVR — publishing creates a zero-day.
Verified leaked secretPVR + rotation request.
Smart-contract bugPVR — on-chain exploitation often immediate.
No PVR + no SECURITY.mdSkip and log. Do no harm.
# PVR (private advisory)
gh api -X POST "/repos/$REPO/security-advisories" \
  -f summary="..." -f severity="..." -F cwe_ids='["CWE-89"]' -f description="..."

Proposed patches for code flaws go to your fork only (private/fix-<slug> branch). Never open a public PR for an unpatched code flaw — link the SHA in the advisory body so the maintainer can cherry-pick.

Required scopes

GH_TOKEN with repo + repository_advisories:write (for PVR).

Rules

  • Do no harm. No safe channel → no publication.
  • Read the code. A scanner hit alone isn't a vulnerability.
  • One report per repo per run; bundle related findings.
  • Skip intentionally vulnerable repos and CTFs.
  • Be deferential — you're offering help, not grading homework.

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