playwright-security-runner
Dynamic security testing of web forms using Playwright browser automation. Sends actual payloads to test for vulnerabilities. REQUIRES USER CONFIRMATION before execution. Use when user wants to "test payloads", "dynamic security test", "exploit testing", "penetration test forms".
When & Why to Use This Skill
The Playwright Security Runner is a specialized Claude skill designed for dynamic security testing and penetration testing of web forms. By leveraging Playwright browser automation, it identifies critical vulnerabilities such as XSS, SQL injection, and authentication bypasses by executing real-world payloads. The tool prioritizes safety with mandatory user confirmations, dry-run modes, and production environment detection, making it an essential asset for developers and security researchers aiming to harden web applications.
Use Cases
- Automated Vulnerability Scanning: Automatically scan web forms for common security flaws like XSS and SQLi during the development lifecycle to ensure robust application security.
- Penetration Testing Support: Assist security professionals in executing and documenting exploit attempts with real browser interactions and automated screenshot evidence.
- Pre-deployment Security Audits: Conduct dynamic security checks on staging or local environments to identify and remediate vulnerabilities before they reach production.
- Bug Bounty Evidence Collection: Streamline the process of verifying vulnerabilities on authorized targets and generating detailed reports with specific payloads and visual proof.
| name | playwright-security-runner |
|---|---|
| description | Dynamic security testing of web forms using Playwright browser automation. Sends actual payloads to test for vulnerabilities. REQUIRES USER CONFIRMATION before execution. Use when user wants to "test payloads", "dynamic security test", "exploit testing", "penetration test forms". |
Playwright Security Runner
Dynamic security testing with real browser automation. This skill sends actual payloads to targets.
CRITICAL: Safety Protocols
This skill sends real requests. ALWAYS get user confirmation first.
Before Running:
- Show the user what payloads will be sent
- Confirm the target is authorized for testing
- Warn if target appears to be production
Confirmation Template:
Security Test Plan
Target: http://localhost:3000/login
Payloads to send:
1. [XSS] <script>alert(1)</script> -> username field
2. [SQLi] ' OR '1'='1 -> password field
This will send real requests to the target.
Potential bounty: $5,000 - $15,000
Proceed? (yes/no)
Quick Start
# Install dependencies
cd ${CLAUDE_PLUGIN_ROOT}/skills/playwright-security-runner && npm install
# Dry run - shows what would be tested (SAFE)
npm --prefix ${CLAUDE_PLUGIN_ROOT}/skills/playwright-security-runner run dev -- \
--url "http://localhost:3000/login" \
--dry-run
# Actually run tests (REQUIRES CONFIRMATION)
npm --prefix ${CLAUDE_PLUGIN_ROOT}/skills/playwright-security-runner run dev -- \
--url "http://localhost:3000/login" \
--test xss,sqli
# Or after building
npm --prefix ${CLAUDE_PLUGIN_ROOT}/skills/playwright-security-runner run build
node ${CLAUDE_PLUGIN_ROOT}/skills/playwright-security-runner/dist/index.js \
--url "http://localhost:3000/login" \
--test xss,sqli
Test Types
| Type | What It Tests | Payloads |
|---|---|---|
xss |
Cross-site scripting | Script tags, event handlers |
sqli |
SQL injection | Quotes, UNION, comments |
auth |
Authentication | Bypass attempts |
Options
| Option | Description |
|---|---|
--url <url> |
Target URL (required) |
--form <selector> |
CSS selector for form (optional) |
--test <types> |
Comma-separated test types |
--dry-run |
Show plan without executing |
--screenshot |
Capture screenshots of results |
--json |
Output as JSON |
--headed |
Run with visible browser |
Safety Features
1. Production Detection
WARNING: Production URL Detected
The target URL appears to be a production system:
https://example.com/login
Security testing against production:
- May cause service disruption
- Could trigger security alerts
- May violate terms of service
Ensure you have authorization to test this target.
2. Dry Run Mode
npm run dev -- --url "http://target.com" --dry-run
Output:
DRY RUN MODE
Target: http://target.com/login
Form: All forms
Tests: xss, sqli
Payloads that would be sent:
[XSS] - Bounty: $500 - $10,000
1. script-tag: <script>alert(1)</script>...
2. img-onerror: <img src=x onerror=alert(1)>...
No requests sent. Remove --dry-run to execute tests.
Output Format
Vulnerability Found
## VULNERABILITY FOUND
**Type**: Reflected XSS
**Severity**: HIGH
**Bounty Estimate**: $2,000 - $10,000
**Target**: http://localhost:3000/search
**Field**: query
**Payload**: <script>alert(1)</script>
**Evidence**:
- Payload reflected in response without encoding
- Alert dialog triggered
**Screenshot**: ./screenshots/xss-001.png
Integration Workflow
- Static Analysis First (safe - no requests)
- Review Findings - Identify high-value targets
- Dry Run - Still safe, just planning
- Get Confirmation - Show payload list to user
- Execute Tests - Now sending payloads
- Document Findings - Screenshots as evidence
Important Notes
- localhost/staging only: Prefer testing against development environments
- Authorization required: Ensure you have permission to test
- Rate limiting: Built-in delays between requests (200ms)
- Evidence collection: Screenshots and logs for bug reports