◄ GASLIGHT KEKKAI ► ARENA ► ninja.ing · offensive assurance

NinjaSploit

The AI-era offensive framework. Recon maps an agent's attack surface (entry points, no entry) → multi-turn, indirect/RAG, and tool-abuse modules break it → a campaign scores it against the OWASP LLM Top 10. Live, against a target you supply. ✓ authorised-use only · benign canaries · SSRF-guarded · your key

I am authorised to security-test this target. NinjaSploit fires benign canary attacks; I accept responsibility for authorised use.

recon is free-ish (8 calls). campaign is heavy (~90 calls on your key).

ninjasploit · idle

// pick a module, configure a target, tick authorised-use, and RUN.

CLI parity: gaslight recon --target agent --url … · gaslight campaign --json report.json. Backend is gated — SSRF-guarded, rate-limited, BYO-key, authorised-use ack, benign-canary payloads. Not an open attack proxy.