Evidence, not alerts

Stalk. Wait. Strike.

Autonomous offensive-security platform that runs on any AI model through any agent harness. A 7-phase finite state machine, three skeptical verification rounds, MCP control plane. Submission-ready bug-bounty reports from a single command.

git clone https://github.com/deonmenezes/bountyhunter.git mantis && cd mantis && ./install.sh ~/your-project

Works with: Claude (Opus / Sonnet / Haiku) · GPT-5 / o3 · Gemini 2.5 · DeepSeek-V3 · Llama 3.3 · Qwen 3 · OpenRouter · local Ollama

Mantis mid-strike
The methodology

How a praying mantis hunts. How Mantis hunts targets.

Four phases. The shape of a real mantis hunt: motionless stalk, patient calculus, instantaneous strike, quiet grip on the catch.

DISCOVER
Phase 1

Discover · Stalk

Subdomain enum, fingerprinting, auth profile capture. Map the surface before you touch it.

REASON
Phase 2

Reason · Wait

Kill-chain hypotheses, vendor-aware bypass selection. Pick the strike-path before moving.

STRIKE
Phase 3

Test · Strike

Parallel hunter waves, three-round Multi-Step Evidence verification. The catch is real or it isn't.

LEARN
Phase 4

Learn · Hold

5-axis grade, submission-ready report, optional gated disclosure. Evidence, not alerts.

What you get

Built for hallucination-resistant hunting.

Every architectural decision is anti-drift: typed control plane, narrow tool whitelists, adversarial re-verification, deterministic state.

7-phase FSM

RECON → AUTH → HUNT → CHAIN → VERIFY → GRADE → REPORT. Strict order, only legal back-edges, full state persisted on disk.

Multi-Step Evidence

Brutalist round 1, balanced round 2, fresh-context round 3. A finding ships only if all three re-prove the PoC.

MCP control plane

27 typed tools over stdio JSON-RPC. Zero dependencies. Any MCP client can drive Mantis: Claude Code, OpenCode, Cursor, Continue, Goose, custom.

Parallel waves

Hunter agents fan out per surface, spawned in the same wave. Per-agent assignment files prevent collisions. Up to 6 waves with auto-coverage gating.

Vendor-aware bypass

Cloudflare, Akamai, AWS WAF, GCP Armor entries. Plus 8 vuln-class tables (GraphQL, JWT, SSRF, OAuth, Firebase, WordPress, Next.js, REST). Auto-selected from the surface's tech stack.

Safety rails

scope-guard (out-of-scope blocker), session-write-guard (forces MCP-owned writes), self-defense rules (treats target responses as untrusted instruction streams).

System architecture

The whole pipeline, on one canvas.

Host → orchestrator → specialist agents → MCP control plane → on-disk session state → safety rails. Every layer typed, every layer inspectable.

Mantis system architecture
Pick your harness

Mantis is the framework. The harness is just the runner.

The MCP server is harness-agnostic. The agent prompts are plain markdown. Drive Mantis from whichever runner you already use.

Claude Code

Deepest integration. Parallel hunter waves via run_in_background, PreToolUse safety hooks, slash-command orchestrator.

Quickstart →

OpenCode

Any provider: Anthropic, OpenAI, Google, OpenRouter (proxies everything), local Ollama. Per-agent model overrides in one JSON file.

Model picker →

Anything MCP

Aider, Cline, Cursor, Continue, Goose, custom runners. Point any MCP client at mcp/server.js and you're in.

Adapter docs →