viperclaw

mcp
Security Audit
Fail
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 7 GitHub stars
Code Fail
  • Hardcoded secret — Potential hardcoded credential in src/viperclaw/analyzers/live_scanner.py
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

AI Pentesting Framework, Bug Bounty MCP Server, SAST/DAST Taint Engine & 12 Offensive Security Skills for Claude Code, Google Antigravity, Codex & Cursor.

README.md
ViperClaw — AI Pentesting Framework, Bug Bounty Automation MCP Server, Multi-Language SAST Taint Analyzer, and 34 Security Audit Skills for Google Antigravity, Claude Code, OpenAI Codex, and Cursor

Latest GitHub Release GitHub Actions CI Status Python 3.10+ Zero Dependencies Model Context Protocol MCP Server 14 Bug Bounty & Pentesting Skills MIT License

ViperClaw (viperclaw / vclaw) upgrades Google Antigravity, Claude Code, OpenAI Codex, and Cursor into a methodology-driven offensive security & bug bounty platform—combining 34 security audit skills, a 13-tool Phase-Gated MCP server, a BrowserOS Neo session bridge, a Stigmergic SQLite Attack Graph, Race / JWT / GraphQL / OOB / SAST / DAST engines, and a deterministic Proof Oracle with zero third-party runtime dependencies.

Website • User Guide • Quickstart • Architecture • CLI & MCP • 34 Skills • Benchmarks


⚡ Why ViperClaw?

Most "AI pentesting wrappers" fail in production: dumping 150+ noisy CLI scanners into an LLM causes context-window rot and instant WAF blocks, while monolithic Docker stacks require 4–6 containers and separate API budgets when you already run a frontier coding agent locally.

ViperClaw pairs 14 methodology-driven playbooks with 10 zero-dependency Python 3.10+ offensive engines and a strict "No Receipt, No Report" Proof Oracle:

Core Subsystem Capability & Implementation
🕸️ Stigmergic Graph Zero-setup SQLite knowledge graph with pheromone priority scoring, decay & Mermaid topology export.
🌐 BrowserOS & Live DAST Bridges signed-in BrowserOS Neo tabs (fetch/XHR hooks, JWTs, storage) + 14 wire-verified HTTP exploit probes.
⚡ Barrier Race Engine Pre-connected socket pool synchronized via threading.Barrier to prove TOCTOU double-spend & coupon overruns.
🔑 JWT & GraphQL Engines HMAC secret cracker, alg:none / kid / RS256→HS256 token forger, plus GraphQL Clairvoyance & batching probes.
📡 OOB & JS Harvester HMAC-tagged blind OOB canary matrix + ephemeral callback listener, plus Webpack / Vite / Next.js lazy-chunk reconstruction.
🔬 Taint SAST & Oracle Multi-hop taint SAST (Python, TS/JS, Go, PHP) + 7-Gate Proof Oracle (WAF Soft-200 filter, BOLA diff, CVSS 3.1, SARIF 2.1.0).

🚀 Quickstart

📘 New to ViperClaw? Read the complete step-by-step Beginner-to-Pro User Guide (docs/README.md).

# 1. Clone & install (<2 seconds, zero external dependencies, Python 3.10+)
git clone https://github.com/zahidoverflow/viperclaw.git
cd viperclaw
pip install -e .

# 2. Install all 34 security audit skills into Antigravity, Claude Code, Codex & Cursor
viperclaw skills install --target all
🔌 Connect the Phase-Gated MCP Server (mcp_config.json / Claude Code / Cursor)

Add viperclaw mcp to your agent's mcp_config.json / .mcp.json (or run claude mcp add viperclaw -- viperclaw mcp):

{
  "mcpServers": {
    "viperclaw": {
      "command": "viperclaw",
      "args": ["mcp"]
    }
  }
}
  • Single-Harness Skill Install: Use --target antigravity, --target claude, --target codex, or --target cursor.
  • Claude Code Plugin Marketplace: Run /plugin marketplace add zahidoverflow/viperclaw then /plugin install viperclaw@viperclaw.

🧬 Architecture

ViperClaw 4-Layer Offensive Architecture — AI Coding Harnesses, 34 Security Audit Skills, 13-Tool Phase-Gated MCP Server, Pure-Python Offensive Analyzers, and 7-Gate Proof Oracle

💻 CLI & MCP Reference

Both viperclaw and the shorthand vclaw alias provide standalone CLI subcommands as well as the JSON-RPC 2.0 stdio MCP bridge:

CLI Command Purpose & Key Flags
viperclaw scan <url> Stateful live HTTP web scanner & wire-verified exploit runner (--db, --report-md, --report-sarif).
viperclaw race <url> Barrier-synchronized concurrent socket burst for TOCTOU race conditions (-c 20, -d, --cookie, --auth).
viperclaw jwt <token> Decode JWT, crack HMAC secrets, forge alg:none / kid / RS256→HS256 tokens & replay (--target-url).
viperclaw graphql <url> Audit GraphQL for introspection bypass, Clairvoyance field suggestions & query batching (--auth, --cookie).
viperclaw oob Generate HMAC-authenticated blind OOB exploit payloads (SSRF, XXE, CMDi, SSTI, SQLi, JNDI) (--callback).
viperclaw browser Output BrowserOS Neo JS runtime hooks (--scripts) or ingest a live browser DOM/storage snapshot (--ingest).
viperclaw taint <path> Multi-language source-to-sink taint SAST for Python, TS/JS, Go & PHP (--db, --output).
viperclaw harvest‑js <path> Webpack / Vite / Next.js lazy-chunk, hidden endpoint & secret harvester (--base-url, --db, --output).
viperclaw fingerprint <har> Caido / Burp .har backend framework & sibling API fingerprinter (--domain, --output).
viperclaw graph Query top pheromone-ranked attack leads & export Mermaid topology (--db, --top-leads, --mermaid).
viperclaw verify [file] Run the 7-Question Triage Gate, CVSS 3.1 calculator & H1 MD / SARIF exporter (--report-md, --report-sarif).
📋 View Copy-Paste CLI Workflow Examples (Live DAST, Race Burst, JWT, GraphQL, OOB, SAST)
# 1. Run a stateful live web scan & export verified HackerOne Markdown + SARIF 2.1.0
viperclaw scan http://127.0.0.1:4280 --db .viperclaw/graph.db --report-md bounty_report.md --report-sarif findings.sarif

# 2. Fire a 20-worker barrier-synchronized TOCTOU race burst against a coupon/transfer endpoint
viperclaw race https://api.target.tld/v1/coupon/redeem -X POST -d "code=VIPER2026" --cookie "session=abc" -c 20

# 3. Crack a JWT's HMAC secret, synthesize forged admin tokens & replay against a live endpoint
viperclaw jwt "eyJhbGciOiJIUzI1NiIs..." --target-url https://api.target.tld/v1/admin

# 4. Scan a GraphQL endpoint for introspection bypass, Clairvoyance field leaks & batching
viperclaw graphql https://api.target.tld/graphql --auth "Bearer eyJ..."

# 5. Audit source code for multi-hop taint flows & reconstruct SPA lazy chunks
viperclaw taint ./target-repo --db .viperclaw/graph.db --output taint_flows.json
viperclaw harvest-js ./bundles --base-url https://app.target.tld/ --db .viperclaw/graph.db
🔌 View All 13 Phase-Gated MCP Server Tools (viperclaw mcp)
MCP Tool Phase Description
viper_live_scan DAST Execute a stateful, authenticated HTTP dynamic scan against a live web target, proving exploits over the wire.
viper_race_burst Exploit Fire a barrier-synchronized concurrent HTTP burst (pre-connected sockets) to prove TOCTOU race conditions.
viper_jwt_attack Exploit Decode JWTs, crack weak HMAC secrets, forge alg:none / kid / key-confusion tokens, and replay against live URLs.
viper_graphql_scan Exploit Audit GraphQL for introspection bypass, Clairvoyance field leaks (Did you mean...?), array batching & BOLA.
viper_oob_probe Exploit Generate HMAC-tagged blind OOB canary payloads (SSRF, XXE, CMDi, SSTI, SQLi, JNDI) and verify callback hits.
viper_browser_ingest BrowserOS Output BrowserOS Neo JS runtime hooks (fetch/XHR/storage) or ingest live browser snapshots into the Attack Graph.
viper_taint_scan SAST Run multi-language source-to-sink taint analysis across Python, TypeScript/JavaScript, Go, and PHP.
viper_harvest_js Recon Reconstruct Webpack/Vite/Next.js lazy chunks and extract endpoints, IDOR/SSRF candidates, GraphQL ops & secrets.
viper_fingerprint_traffic Recon Fingerprint backend frameworks, WAFs, sibling API hosts, and error envelopes from .har captures or HTTP headers.
viper_graph_upsert Graph Upsert hosts, endpoints, lazy chunks, taint sinks, or findings into the Stigmergic SQLite Attack Graph.
viper_graph_leads Graph Query top pheromone-ranked attack leads and export a live Mermaid topology diagram.
viper_diff_oracle Verify Compare baseline vs. mutated/cross-tenant HTTP responses to confirm BOLA/IDOR and filter WAF Soft-200 blocks.
viper_verify_and_report Report Run the 7-Question Triage Gate, calculate exact CVSS v3.1 scores, and generate HackerOne/Bugcrowd Markdown.

🎯 Offensive Skill Playbooks (34 Skills)

Each skill in skills/ encodes a senior security researcher's hypothesis tree, bypass ladder, and stopping conditions across 5 operational domains:

📚 View Full 34-Skill Directory & Methodology Breakdown
# Skill Playbook Security Audit & Remediation Focus
01 viper‑ai‑llm‑redteam AI, LLM & MCP Application Security Auditor (OWASP LLM Top 10).
02 viper‑auth‑oauth‑jwt Authentication, OAuth 2.
03 viper‑bola‑idor Object & Function-Level Authorization (BOLA / BFLA) Audit Skill.
04 viper‑browseros‑agent BrowserOS Neo Live Application Inspection & DOM Security Audit Skill.
05 viper‑cicd‑hardening CI/CD Pipeline & GitHub Actions Security Auditor (modeled on Zizmor, Actionlint & StepSecurity).
06 viper‑container‑k8s‑audit Docker Container & Kubernetes Manifest Security Auditor (modeled on Trivy, Kubescape & CIS Benchmarks).
07 viper‑crypto‑tls‑audit Cryptography, Key Management & Password Hashing Auditor (modeled on Trail of Bits & OWASP Cryptographic Storage).
08 viper‑csrf‑samesite‑audit Cross-Site Request Forgery (CSRF) Defense & State-Change Auditor (OWASP ASVS V4).
09 viper‑dependency‑sbom Software Supply Chain, Lockfile Integrity & Dependency Vulnerability Auditor (modeled on OSV-Scanner, Pip-Audit & Dependabot).
10 viper‑deserialization‑parser Safe Deserialization & XML/YAML/JSON Parser Auditor (OWASP Deserialization Cheat Sheet).
11 viper‑file‑upload‑storage File Upload, Path Traversal & Object Storage Security Auditor (OWASP File Upload Cheat Sheet).
12 viper‑graphql‑ws GraphQL Schema & WebSocket Security Auditor.
13 viper‑headers‑csp‑cors HTTP Security Headers, Content-Security-Policy (CSP) & CORS Configuration Auditor (modeled on Mozilla Observatory & OWASP Secure Headers).
14 viper‑iac‑terraform‑audit Infrastructure-as-Code (Terraform, CloudFormation & Pulumi) Security Auditor (modeled on Checkov & tfsec).
15 viper‑injection‑oracle Input Validation & Query Parameterization Auditor.
16 viper‑js‑harvester Client-Side Bundle & SPA Manifest Auditor.
17 viper‑logging‑pii‑compliance Security Logging, PII Redaction & Error Sanitization Auditor (OWASP Logging Cheat Sheet & ASVS V7).
18 viper‑mobile‑api‑config Mobile App Static Configuration & API Security Auditor (modeled on OWASP MASVS & MobSF).
19 viper‑multitenant‑isolation Multi-Tenant SaaS Data Isolation & Row-Level Security (RLS) Auditor.
20 viper‑oauth‑oidc‑compliance OAuth 2.
21 viper‑openapi‑rest‑schema OpenAPI / Swagger Schema, Mass Assignment & REST API Security Auditor (OWASP API3 & API4).
22 viper‑orchestrator Master 5-Phase Security Audit & Architecture Review Orchestrator.
23 viper‑proof‑judge Evidence Quality Gate, CVSS v3.
24 viper‑proxy‑fingerprint HTTP Stack, Header & Error-Envelope Auditor.
25 viper‑race‑logic Concurrency, Idempotency & Business Logic Auditor.
26 viper‑rate‑limit‑dos‑resilience Application Rate Limiting, ReDoS & Resource Exhaustion Auditor (OWASP API4:2023).
27 viper‑remediation‑patch‑verifier Security Patch Synthesizer & Regression Test Generator.
28 viper‑secrets‑audit Hardcoded Secret, Credential Entropy & Git History Leak Auditor (modeled on Gitleaks & TruffleHog).
29 viper‑session‑cookie‑audit Session Lifecycle, Cookie Hardening & Token Invalidation Auditor (OWASP ASVS V3).
30 viper‑smart‑contract‑review Smart Contract Static Security Auditor (modeled on Slither, Trail of Bits & OpenZeppelin checklists).
31 viper‑ssrf‑cloud Outbound Request (SSRF) & Cloud Egress Auditor.
32 viper‑taint‑sast White-Box Multi-Language Source-to-Sink Taint Analysis Skill.
33 viper‑waf‑evasion‑framing WAF Configuration, Parser Normalization & Filter Consistency Auditor.
34 viper‑webhook‑edr‑hmac Inbound & Outbound Webhook Security Auditor.

🏆 Benchmarks & Comparison

ViperClaw is continuously validated against Damn Vulnerable Web Application (DVWA) and synthetic TOCTOU/JWT/GraphQL/OOB labs across both static source auditing and live wire exploitation:

  • White-Box Taint SAST (viperclaw taint): 19 / 19 (100%) vulnerable modules detected via transitive taint tracking, with 0 False Positives on hardened impossible.php targets (PDO::prepare, intval, is_numeric, in_array, htmlspecialchars).
  • Black-Box Live Exploitation (viperclaw scan + race + jwt + graphql + oob): 14 / 14 (100%) DVWA modules plus barrier-synchronized TOCTOU race windows, HMAC JWT cracking, GraphQL Clairvoyance, and OOB callbacks verified over the wire.
🔬 View All 14 Wire-Verified DVWA Exploitation Proofs & CVSS v3.1 Vectors
# Severity Vulnerability Module Verified Wire Proof / Exploit Vector
01 HIGH (8.8) OS Command Injection ip=127.0.0.1;id → Verified uid=33(www-data) execution
02 HIGH (8.1) UNION SQL Injection ' UNION SELECT user,password FROM users# → Credential hash dump
03 MED (6.5) Boolean Blind SQLi True (1' AND 1=1#) vs. False (1' AND 1=2#) differential oracle
04 HIGH (8.1) SQLi Auth Bypass username=admin' OR '1'='1' LIMIT 1# → Authenticated admin session
05 HIGH (7.1) Local File Inclusion ?page=/etc/passwd → Verified root:x:0:0:root:/root:/bin/bash
06 CRIT (9.9) Unrestricted Upload RCE Multipart PHP upload → Verified RCE at /hackable/uploads/viperclaw_poc.php
07 MED (6.1) Reflected XSS Unescaped HTML payload reflection verified in DOM response
08 MED (5.4) Stored XSS (Guestbook) Persistent payload stored in mtxMessage and rendered to viewers
09 HIGH (8.2) Predictable Session ID Sequential dvwaSession $+1$ increment proven across consecutive requests
10 HIGH (8.8) CSRF Password Reset State-changing GET password reset verified without anti-CSRF token
11 MED (6.5) CAPTCHA Workflow Bypass Direct step=2 state-machine jump verified via auth oracle & restored
12 MED (5.4) CSP Bypass & Injection External script injection via permissive Content-Security-Policy & include
13 MED (6.5) JS Token Reversal Reversed md5(rot13("success")) client-side token & verified acceptance
14 MED (6.1) DOM-Based XSS Unescaped location.href sink into document.write() verified in browser
📊 View 2026 AI Pentesting Framework Comparison Matrix (ViperClaw v2.0 vs. Strix, PentAGI, HexStrike, Vulnhuntr)
Capability / Architecture 🐍 ViperClaw v2.0 Strix / Shannon PentAGI / RedAmon HexStrike AI Vulnhuntr
Architecture Pattern 34 Skills + 13 MCP + Engine Containerized Agent Docker + Neo4j Stack 150-Tool MCP Python AST CLI
Native Inside Antigravity / Claude / Codex / Cursor ✅ Native ⚠️ CLI Wrapper ⚠️ Partial ✅ MCP Only ❌ Standalone
Signed-In BrowserOS Neo DOM & Fetch Bridge ✅ Built-in ⚠️ Isolated Container ❌ ❌ ❌
Barrier-Synchronized Socket Race Engine ✅ Built-in ❌ ❌ ❌ ❌
JWT Cracker + Forger & GraphQL Clairvoyance ✅ Built-in ⚠️ Partial ❌ ⚠️ External CLI ❌
Cryptographic OOB Canary & Ephemeral Listener ✅ Built-in ✅ OOB ⚠️ Partial ❌ ❌
Zero-Setup SQLite Attack Graph + Mermaid ✅ Built-in ❌ ⚠️ Needs Neo4j/PG ❌ ❌
SPA Lazy-Chunk Reconstruction (Webpack/Vite/Next) ✅ Built-in ⚠️ Partial ❌ ❌ ❌
Multi-Language White-Box Taint SAST ✅ Py / TS / Go / PHP ✅ Dynamic + Code ❌ Black-Box ❌ ⚠️ Py 3.10 Only
Deterministic WAF Soft-200 & BOLA Diff Oracle ✅ Built-in ✅ PoC Validation ⚠️ LLM-Judged ❌ Raw Stdout ❌ Static Only
Runtime Dependencies ✅ 0 (Pure Stdlib) Docker + Node/LLM Docker + DB + Kali 150+ CLI Tools Python 3.10 + jedi
📂 View Repository Structure & Package Layout
viperclaw/
├── .claude-plugin/               # Claude Code Marketplace & Plugin manifests
├── .github/workflows/ci.yml      # GitHub Actions CI matrix (Python 3.10, 3.11, 3.12)
├── assets/                       # Custom Light-Theme SVG Banner & Architecture Diagrams
├── docs/README.md                # Beginner-to-Pro Complete Usage & Best-Utilization Guide
├── logo.png                      # Lightweight 512x512 transparent square emblem
├── skills/                       # 14 Methodology-Driven Offensive Skills (SKILL.md)
├── src/viperclaw/                # Zero-Dependency Python 3.10+ Core Package
│   ├── cli.py                    # CLI entrypoints (`viperclaw` & `vclaw`)
│   ├── mcp_server.py             # Phase-Gated 13-Tool JSON-RPC 2.0 stdio MCP Server
│   ├── core/                     # Stigmergic Attack Graph, Proof Oracle & SARIF Reporter
│   └── analyzers/                # Live DAST, Race, JWT, GraphQL, OOB, BrowserOS, Taint, JS & HAR
├── tests/test_viperclaw.py       # 11 End-to-End & Adversarial QA Regression Suites
├── CITATION.cff                  # Academic & Research Citation Metadata
├── glama.json                    # MCP Registry Manifest
├── llms.txt                      # AI / LLM Context Discovery Index
└── pyproject.toml                # PEP 621 Zero-Dependency Package Configuration
❓ Frequently Asked Questions (FAQ)
  • Why use ViperClaw instead of a standalone Docker pentest agent?
    If you already use Google Antigravity, Claude Code, OpenAI Codex, or Cursor, spinning up a separate 5-container Docker stack with its own LLM API billing is redundant. ViperClaw installs in 2 seconds with zero dependencies and turns your existing coding agent into an evidence-gated offensive security platform.
  • How does ViperClaw prevent AI hallucinations and WAF "Soft-200" false positives?
    Generic AI wrappers often report 200 OK HTML challenge pages from Cloudflare, Akamai, Imperva, or CloudFront as vulnerabilities. ViperClaw's ProofOracle filters WAF soft-blocks, diffs structural JSON keys for BOLA/IDOR, and enforces a strict 7-Question Pre-Submission Gate ("No Receipt, No Report").
  • How do I run the regression test suite?
    Run pytest -v from the repository root to execute all 11 end-to-end and adversarial QA test suites.

📜 Documentation, Security & License

Reviews (0)

No results found