Example: Greybox scan, multiple domains

User: "Test app.corp.com, api.corp.com, admin.corp.com as admin@corp.com / <pw>."

Confirm scope + tier of account; ask for second (low-priv) account for IDOR.
Spawn 3 sub-agents via the Task tool, one per domain. Each executes workflows/greybox_authenticated.md.
Each sub-agent:
- Runs Phase 0 recon (parallel inside itself).
- Logs in via Playwright, persists storageState.json in its own directory.
- Runs authed crawl and vuln tests.
- Writes results/<domain>/output.json.
Parent agent waits on all sub-agents.
Bash → merge per-domain outputs into results/aggregate.json.
Re-verify all Critical/High across all domains.
Return executive summary + path to aggregate output.

Max 5 parallel sub-agents; here we use 3 (domain count).
Each sub-agent caps at 5 req/s.
Auth is sequential inside each sub-agent; crawl is parallel across sub-agents.

If the operator supplies both admin@corp.com and user@corp.com:

Run two sub-agents per target (one per account).
The IDOR comparator then diffs object-ID reachability between the two storage states.

Read credentials from env (DAST_USER, DAST_PASS) — never inline in tool calls.
Scrub logs: grep -r "$DAST_PASS" results/ must return nothing before delivery.

Third-party checks

1 alert3mo3 checks · Risk SAFE

Gen Agent Trust Hub3mo
The skill provides a comprehensive framework for automated dynamic application security testing (DAST). It utilizes Playwright for browser automation and integrates various security tools like Nuclei and SQLMap. While the skill includes numerous attack patterns and destructive payloads, these are documented as reference material and test cases to be applied against target systems. The skill follows security best practices for credential management and scoping.
Socket3mo
2 alerts: gptSecurity
Snyk3mo
Risk: CRITICAL · 2 issues

/dast-automation