Auditing Cors Policy

Auditing CORS Policy

Overview

CORS misconfiguration is one of the most common middle-severity findings in web bug bounties. The browser-enforced rules are subtle, the failure modes are silent (the wrong cors response just works until an attacker weaponizes it), and the "fix" engineers reach for — Access-Control-Allow-Origin: * — opens the very class of attacks CORS was meant to prevent when paired with credentials.

This skill probes each common CORS misconfiguration with synthetic Origin headers and grades the response.

When the skill produces findings

Prerequisites

• Python 3.9+ (requests library)
• Authorization for non-local targets (../analyzing-tls-config/references/AUTHORIZATION.md)

Instructions

Step 1 — Confirm Authorization

"Do you have authorization to perform CORS testing on this target?
 I need confirmation before proceeding."

Step 2 — Run the scanner

python3 ${CLAUDE_PLUGIN_ROOT}/skills/auditing-cors-policy/scripts/audit_cors.py \
    https://api.example.com/endpoint \
    --authorized

Options:

Usage: audit_cors.py URL [OPTIONS]

Options:
  --authorized       Attest authorization for non-local targets (required)
  --output FILE      Write findings to FILE
  --format FMT       json | jsonl | markdown (default: markdown)
  --min-severity SEV (default: info)
  --timeout SECS     Per-probe timeout (default: 10)
  --method METHOD    HTTP method for the main probe (default: GET)

The scanner sends multiple probes per target:

1. Baseline request with no Origin header
2. Probe with safe Origin (https://allowed-origin.example.com)
3. Probe with attacker Origin (https://attacker.example)
4. Probe with subdomain-bypass Origin
5. Probe with Origin:null
6. OPTIONS preflight with Access-Control-Request-Headers / Method

For each, it records the response's CORS headers and grades against the threshold table above.

Step 3 — Interpret findings

CRITICAL = credential-stealing chain available; ship same-day fix. HIGH = arbitrary cross-origin read of public-but-sensitive content; ship within sprint. MEDIUM/LOW = posture hardening; backlog.

Step 4 — Cross-skill chaining

If CORS findings land alongside auth findings (skill #20 confirming- pentest-authorization would have caught the auth side at engagement scope), suggest authentication-validator plugin for full session- handling audit.

Examples

Example 1 — Reflected-origin bug bounty triage

User: "Bug bounty submission claims CORS bypass on /api/profile."

python3 ${CLAUDE_PLUGIN_ROOT}/skills/auditing-cors-policy/scripts/audit_cors.py \
    https://api.example.com/profile \
    --authorized \
    --format json | jq '.[] | select(.severity == "critical" or .severity == "high")'

If the reflected-origin probe + Allow-Credentials:true both fire, the submission is valid; pay the bounty and ship the fix from PLAYBOOK.md.

Example 2 — Pre-launch CORS sweep on a new API gateway

User: "We're rolling out a new gateway — audit every endpoint's CORS posture."

for ENDPOINT in $(cat endpoints.txt); do
  python3 ${CLAUDE_PLUGIN_ROOT}/skills/auditing-cors-policy/scripts/audit_cors.py \
      "$ENDPOINT" --authorized --min-severity high --format jsonl >> cors-audit.jsonl
done
jq -s 'group_by(.severity) | map({sev: .[0].severity, count: length})' cors-audit.jsonl

Example 3 — CI guard against CORS regression

Drop into deploy gate:

- name: CORS posture gate
  run: |
    python3 plugins/security/penetration-tester/skills/auditing-cors-policy/scripts/audit_cors.py \
        https://staging-api.example.com/auth \
        --authorized \
        --min-severity high

Exit 1 fails the deploy if any new high/critical lands.

Output

JSON / JSONL / Markdown per lib/report.py. Exit 0 clean, 1 high/ critical, 2 error.

Error Handling

• Target returns no CORS headers at all → INFO finding "CORS not configured" with note that this is fine for non-cross-origin endpoints.
• Preflight fails with 405 → MEDIUM finding suggesting OPTIONS handler missing.
• Connection error → exit 2 with underlying error.

Resources

• references/THEORY.md — How CORS works, why each finding matters
• references/PLAYBOOK.md — Allow-list config templates per server type / framework (nginx, Express, Spring, FastAPI, Rails)
• ../analyzing-tls-config/references/AUTHORIZATION.md — Active-scan authorization pattern

Source: jeremylongshore/claude-code-plugins-plus-skills → skills/.curated/auditing-cors-policy/SKILL.md

Also appears in: jeremylongshore/claude-code-plugins-plus-skills/plugins/security/penetration-tester/skills/auditing-cors-policy/SKILL.md