Create your account
Sign up at rasec.app/signup. The free tier is fully functional — no credit card required. You get one concurrent hunt session, surface-level findings, and 30-day data retention.
Documentation
Everything you need to run your first overnight hunt, understand findings, and export report-ready PoCs.
📖 In this guide
Quick start
Sign up at rasec.app/signup. The free tier is fully functional — no credit card required. You get one concurrent hunt session, surface-level findings, and 30-day data retention.
Before launching, set your scope. Paste the program's in-scope domains and CIDR ranges. RaSEC Hunt gates every agent action against this boundary mathematically before execution.
# Example scope configuration
in_scope:
- "*.example.com"
- "api.example.com"
- "10.0.0.0/8" # internal range if permitted
out_of_scope:
- "support.example.com"
- "status.example.com"Hit "Start Hunt". The agent begins recon: subdomain enumeration, endpoint discovery, JavaScript analysis, and coordinated vulnerability testing. All visible in the real-time mission log.
Every finding has an explicit state: Pending (AI heuristic flagged), Validated (deterministic reproduction passed), or Rejected (false positive). Pro and Elite tiers include the curl command for manual verification.
# Finding lifecycle
Pending → AI heuristic detected an anomaly
Validated → Deterministic reproduction confirmed
Rejected → False positive (agent discarded)
# You only act on: ValidatedSelect the validated finding and click "Generate PoC". The report formats itself to your chosen template: HackerOne, Bugcrowd, Intigriti, or custom. Copy the output and submit.
Common questions
Every agent action (DNS lookup, HTTP request, payload test) is checked against your defined scope before execution. If the target is out of scope, the action is blocked at the gate, not after the fact.
When the AI flags a potential vulnerability, it runs a second, independent reproduction attempt with a non-AI approach. If that also confirms the bug, the finding is marked Validated. If not, it is Rejected. You only see confirmed findings.
Yes — co-pilot mode (Pro and Elite) lets you intervene at any point. Redirect the agent to a specific endpoint, pause a module, or change priority mid-hunt. Full keyboard control.
Surface-level issues: missing security headers, misconfigured CORS, exposed debug endpoints, directory listing, and open redirect indicators. Serious bugs (IDOR, auth bypass) require Pro.
No. We do not train on hunt data. We do not log raw HTTP scan traffic. Free tier data is auto-purged after 30 days. Pro includes no-log mode. Elite supports Local Proxy Mode (zero cloud scan traffic).