← ClaudeAtlas

sec-scanlisted

Self-contained security audit for web properties: response-header hardening, CSP policy linting, TLS config, exposed sensitive files, secret leak scan, port/OS detection, and version-leak checks. Use when asked to security-scan / audit / pentest-lite a site or VPS before launch or after a security report arrives. Vendored into fable; routed automatically by scripts/route.py for security-audit tasks.
romangalaxys10-spec/fable · ★ 5 · AI & Automation · score 62
Install: claude install-skill romangalaxys10-spec/fable
# sec-scan — self-contained security audit (vendored in fable) Re-audit a web property (and optionally the VPS it lives on and the repo behind it) without paid scanners. The single tool is `secscan.py`, a stdlib-only Python 3.10+ script in this folder. ## When to use - Before launch, or right after a security report (nmap/nuclei/zap/nikto/ spiderfoot) arrives and you want to confirm or reproduce the key findings in-house. - When the team asks for a quick "is our site still safe?" pass. - Weekly-ish hygiene check: run it, diff the HTML report against the last one. ## How to run Single command, no pip installs: ```bash python3 <plugin-root>/vendor/sec-scan/secscan.py <target> [--vps <ip>] [--repo <path>] \ [--port-sweep top1000|full] [--compliance] [--out report.html] ``` `<target>` may be a bare hostname or an http(s) URL. Exit code is 0 when there are no Critical/High findings, 1 otherwise — so it is cron- and CI- friendly. ## What each check does - **A. HTTP header audit** — HSTS, Server/version leaks, X-Powered-By/AspNet/ Generator/Via, X-Content-Type-Options, clickjacking, Referrer/Permissions- Policy, cookie Secure/HttpOnly, open-redirect probes (3–4 callback paths). - **B. CSP lint (ZAP-equivalent rules)** — missing directives with no `default-src` fallback, wildcards, `unsafe-inline`/`unsafe-eval`, worker/child-src gaps. - **C. TLS audit** — cert expiry (<14d High, <30d Medium), TLS 1.0/1.1 enabled, self-signed cert. - **D. Sensitive-file prob