TyrusRC/praetor
Agentic pentest & red-team harness via Burp Suite MCP toolkit
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit TyrusRC/praetor/.claude ./rig-praetor # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ curl -fsSL --create-dirs -o .claude/agents/auth-payment-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/auth-payment-agent.md $ curl -fsSL --create-dirs -o .claude/agents/auth-tester.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/auth-tester.md $ curl -fsSL --create-dirs -o .claude/agents/browser-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/browser-agent.md $ curl -fsSL --create-dirs -o .claude/agents/cloud-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/cloud-agent.md $ curl -fsSL --create-dirs -o .claude/agents/desktop-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/desktop-agent.md $ curl -fsSL --create-dirs -o .claude/agents/finding-verifier.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/finding-verifier.md $ curl -fsSL --create-dirs -o .claude/agents/fuzz-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/fuzz-agent.md $ curl -fsSL --create-dirs -o .claude/agents/grow-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/grow-agent.md $ curl -fsSL --create-dirs -o .claude/agents/js-analyst.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/js-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/llm-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/llm-agent.md $ curl -fsSL --create-dirs -o .claude/agents/mobile-dynamic-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/mobile-dynamic-agent.md $ curl -fsSL --create-dirs -o .claude/agents/network-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/network-agent.md $ curl -fsSL --create-dirs -o .claude/agents/payload-crafter.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/payload-crafter.md $ curl -fsSL --create-dirs -o .claude/agents/pentest-commander.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/pentest-commander.md $ curl -fsSL --create-dirs -o .claude/agents/recon-agent.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/recon-agent.md $ curl -fsSL --create-dirs -o .claude/agents/redteam-commander.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/redteam-commander.md $ curl -fsSL --create-dirs -o .claude/agents/screenshot-triage.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/screenshot-triage.md $ curl -fsSL --create-dirs -o .claude/agents/ste-writer.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/ste-writer.md $ curl -fsSL --create-dirs -o .claude/agents/vuln-scanner.md https://raw.githubusercontent.com/TyrusRC/praetor/main/.claude/agents/vuln-scanner.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"hooks": {
"PreToolUse": [
{
"matcher": "mcp__praetor__record_retest|mcp__praetor__mark_finding_false_positive",
"hooks": [
{
"type": "command",
"command": "python3 \"$CLAUDE_PROJECT_DIR/.claude/hooks/finding-guard.py\""
}
]
}
]
}
} Subagents (19)
| auth-payment-agent | Deep-dive OAuth/OIDC, WebAuthn/FIDO2/passkeys, Apple/Google/Samsung Pay, IAP receipt validation, 3DS 2.x bypass, SCA exemption abuse, recovery downgrades. $5k-$50k bug class. |
| auth-tester | Test authorization and access control across endpoints with ≥2 auth states. Returns IDOR / BFLA / auth-bypass findings. |
| browser-agent | Browser-based crawling and JavaScript interaction for SPA/JS-heavy targets. Populates Burp Proxy history with dynamic routes and XHR/API calls. |
| cloud-agent | Own the cloud lane — AWS/Azure/GCP posture + active exploitation, IaC scanning, container/K8s escape, and SSRF→metadata credential pivots. Returns cloud findings with the misconfig→impact chain, not r |
| desktop-agent | Worker agent for desktop application security testing — Electron / Tauri / WebView2 binary inspection + IPC fuzzing + auto-update MITM. 1-per-binary, no recursion. |
| finding-verifier model: haiku | Re-verify suspected/confirmed findings and investigate anomalies. Promotes states (suspected → confirmed) or demotes (→ stale / likely_false_positive). |
| fuzz-agent | Discover hidden directories and files using tech-aware SecLists slicing. Replaces spray fuzzing with surgical wordlists. |
| grow-agent | Session orchestrator for one domain. Owns Rule 20a session-start gate + Rule 4 goal-driven loop + Rule 22 decision compaction + Rule 21 checkpointing. Promotes confirmed cross-target patterns into KB/ |
| js-analyst model: haiku | Deep JavaScript analysis — secrets, DOM sinks, hidden API endpoints. Returns enriched JS intel for the orchestrator. |
| llm-agent | Test the LLM/AI layer of a captured endpoint (OWASP LLM Top 10) — prompt injection, jailbreak/encoding bypass, system-prompt leakage, improper output handling into a sink, sensitive-data disclosure, e |
| mobile-dynamic-agent | Drive Frida (iOS+Android) and adb (Android) on operator's host. Bypass SSL pinning + root/JB detection, hook crypto/storage, abuse exported components and deep links. Dynamic-only; no static decompile |
| network-agent | Own the network / Active Directory lane — discovery, service enum, credential capture→crack→reuse, lateral movement leads. Returns network-lane findings (ATT&CK operator-log ids, not Burp indices) + a |
| payload-crafter | Craft bypass payloads when standard attacks are blocked by WAF/filters. Returns working bypass or "filter too strong" with evidence. |
| pentest-commander model: opus | Pentest engagement lead. Owns research, planning, multi-domain grow-agent dispatch, cross-target synthesis, and the findings report. Coverage-driven — success is full WSTG/OWASP coverage with every fi |
| recon-agent model: haiku | Map a target's attack surface — endpoints, tech stack, sensitive files, hidden parameters. Returns enriched intel for the orchestrator. |
| redteam-commander model: opus | Red team engagement lead. Owns research, planning, multi-domain grow-agent dispatch, cross-target synthesis, and the attack-narrative report. Objective-driven — success is the stated objective reached |
| screenshot-triage model: haiku | Read ONE saved screenshot and return a compact list of leads (forms, endpoints, error strings, auth markers) — cheap OCR-first triage, keeps vision tokens out of the main agent's context. |
| ste-writer model: haiku | Rewrite text into ASD-STE100 Simplified Technical English — short, unambiguous, controlled English for reports, tool descriptions, agent-to-agent messages, and error messages. Returns the rewritten te |
| vuln-scanner | Test ONE vulnerability category on assigned non-overlapping endpoints. Returns findings + anomalies for orchestrator review. |
Hooks (4)
| event | matcher | runs |
|---|---|---|
| PreToolUse | mcp__praetor__record_retest|mcp__praetor__mark_finding_false_positive | python3 "$CLAUDE_PROJECT_DIR/.claude/hooks/finding-guard.py" |
| PostToolUse | mcp__praetor__(save_finding|record_retest|lock_finding|lock_findings|unlock_finding|mark_finding_false_positive|assess_f | python3 "$CLAUDE_PROJECT_DIR/.claude/hooks/activity-log.py" |
| SessionStart | * | python3 "$CLAUDE_PROJECT_DIR/.claude/hooks/session-start.py" |
| PreCompact | * | python3 "$CLAUDE_PROJECT_DIR/.claude/hooks/pre-compact.py" |
Slash commands (2)
/checklist/exploit
Cursor rules (1)
praetor · always
Similar rigs
Mikacr1138/claude-bug-bounty
Enable efficient bug bounty hunting across Web2 and Web3 with a tool that supports full recon to detailed reporting.
Orchestrator 1.7k tok · · Q0
0xSteph/pentest-ai-agents
Turn Claude Code into your offensive security research assistant. Specialized AI subagents for authorized penetration testing plan engagements, analyze recon, research exploits, build detections, audit STIGs, and write reports.
Orchestrator 3.1k tok · · Q0
mryll/skills
Portable agent skills for AI coding agents — works with Claude Code, OpenAI Codex CLI, Gemini CLI, Cursor & GitHub Copilot: code review, vertical-slice architecture, Go testing strategy, low cognitive-complexity & AGENTS.md generation. Mark
Skill Collector 1.9k tok · · Q0
voxpelli/vp-claude
Claude Code plugin marketplace + plugin that turns Basic Memory into an actively maintained knowledge graph
Orchestrator 10.7k tok · · Q8