/pentest

Security Testing Workflow

Run a structured security assessment. This command walks through authorization, scope selection, scanning, and reporting.

Step 1: Authorization Check

Before scanning anything, confirm authorization:

• If the target is a URL: ask the user to confirm they own it or have written permission to test it.
• If the target is local code/dependencies: confirm it is the user's own project.
• Do not proceed without explicit authorization.

Step 2: Determine Scope

Ask the user what they want to test:

1. Web application (URL) -- security headers, SSL, exposed endpoints, CORS
2. Dependencies (project directory) -- npm/pip vulnerability audit
3. Source code (directory) -- static analysis for secrets, injection, etc.
4. Full audit -- all of the above

Step 3: Run Scanners

Based on the selected scope, run the appropriate scripts from the plugin:

Web Application Scan

python3 ${CLAUDE_PLUGIN_ROOT}/skills/performing-penetration-testing/scripts/security_scanner.py TARGET_URL --verbose

Dependency Audit

python3 ${CLAUDE_PLUGIN_ROOT}/skills/performing-penetration-testing/scripts/dependency_auditor.py TARGET_DIR --verbose

Code Security Scan

python3 ${CLAUDE_PLUGIN_ROOT}/skills/performing-penetration-testing/scripts/code_security_scanner.py TARGET_DIR --verbose

Save JSON reports for any scan that finds critical or high issues:

python3 SCANNER --output /tmp/security-report-$(date +%Y%m%d).json

Step 4: Present Findings

Summarize results for the user:

1. Summary table -- total findings by severity across all scanners
2. Critical/High findings -- detail each one with the risk and impact
3. Remediation priorities -- ordered list of what to fix first

Step 5: Suggest Remediations

For each critical and high finding:

1. Explain the vulnerability in plain language
2. Provide the specific fix (reference REMEDIATION_PLAYBOOK.md)
3. Show how to verify the fix

Offer to apply code fixes directly for code-level findings.

Step 6: Generate Report

If the user wants a saved report, combine all findings into a single JSON file:

# Reports are saved via the --output flag on each scanner

Safety Rules

• Never run scans against unauthorized targets
• All scanners use safe, non-destructive techniques (GET requests, static analysis)
• No exploit payloads are sent to targets
• No data is exfiltrated or modified