← ALL DOCS

AI SECURITY AGENTS

Eight specialised agents share the crawl and attack your app from different angles. Stealth agents go first, aggressive ones follow, and results are deduplicated into one finding list.

The agent team

Agent Focus Style
Web Reconnaissance Scout Maps surface area via crawling, fingerprinting, and subdomain enumeration. Balanced
Injection Vulnerability Hunter SQLi, XSS, command injection, template injection. Aggressive
Authentication Analyzer Auth bypass, weak credentials, session fixation, JWT weaknesses. Aggressive
API Security Prober REST and GraphQL endpoints for IDOR and mass assignment. Balanced
Configuration Auditor Headers, TLS, exposed config files, directory listing. Stealth
Business Logic Tester Flow bypass, privilege escalation, race conditions. Balanced
SSRF & Path Traversal Crawler Server-side request forgery and path traversal. Stealth
Compliance Risk Assessor GDPR, PCI-DSS, cookie policy, and data-exposure risks. Stealth

Each agent card on the Agents page lists its specialties, tactics, average duration, and average findings.

How a run is coordinated

  1. Select agents — the enabled set is used for every scan in your organization.
  2. Parallel execution — stealth agents run first to complete reconnaissance without tripping defences; aggressive agents follow.
  3. Coordinated results — findings are deduplicated and correlated across agents so one root cause produces one finding with related findings linked.

Enabling and disabling

Toggle any agent on its card, or use Enable All / Disable All. The header shows the estimated total duration and expected findings for the current selection.

Editing an agent profile

Open an agent to see its JS Profile — the script that drives its behaviour. You can edit and save it (an Unsaved changes indicator appears while editing). Custom scan scenarios in JavaScript are a Pro and Lifetime feature; keep a copy of the default before making large changes.

COMMON QUESTIONS

01
Do I have to enable agents manually?
No. All eight are enabled for a new organization. Use the Agents page to switch individual agents off — for example disable the aggressive injection and auth agents on a rate-limited production app.
Q&A
02
Does disabling an agent make scans faster?
Yes. The Agents page shows an estimated total duration and expected findings for the enabled set, and updates as you toggle agents.
Q&A

SEE IT ON YOUR OWN APP

Docs explain the product. A scan shows you what it finds on your deployed URL — keys, RLS, auth, and API gaps in under 60 seconds.

14-day free trial · No credit card · Cancel anytime

START FREE SCAN