# ⚔️ The Analyzer v1.0 — Autonomous Bug Bounty Engine **Authorized Bug Bounty Use Only** — Drjonesxxx / Indianaholmes ## One Command ```bash ./analyzer https://target.com ``` Or for deep scan: ```bash ./analyzer https://target.com deep ``` ## Modes | Mode | Command | What it does | |------|---------|-------------| | **Quick** (default) | `./analyzer https://x.com` | Recon → Ollama picks top vectors → reports | | **Deep** | `./analyzer https://x.com deep` | Recon → ALL 20 vectors → full report | | **Custom** | `./analyzer https://x.com custom` | Pick your own vector numbers | | **Interactive** | `./analyzer` | Menu-driven mode | ## How it Works ``` ┌─────────────┐ Target URL ────────▶│ RECON │───▶ HTTP headers, tech detection, │ ENGINE │ endpoints, forms, cookies └──────┬──────┘ │ ▼ ┌─────────────┐ │ OLLAMA │───▶ Analyzes recon data │ BRAIN │ Picks best 5-10 vectors └──────┬──────┘ │ ┌──────────────┼──────────────┐ ▼ ▼ ▼ ┌──────────┐ ┌──────────┐ ┌──────────┐ │ SQLi │ │ XSS │ │ LFI │ ... 20 vectors │ Module │ │ Module │ │ Module │ └──────────┘ └──────────┘ └──────────┘ │ │ │ ▼ ▼ ▼ ┌─────────────┐ │ REPORT │───▶ Markdown report │ ENGINE │ HTML report (beautiful) └─────────────┘ CLI summary ``` ## 20 Attack Vectors | # | Vector | Severity | Detects | |----|--------|----------|---------| | 01 | SQL Injection | CRITICAL | SQLi (error, blind, time) via sqlmap + manual | | 02 | XSS | HIGH | Reflected, DOM-based | | 03 | LFI/RFI | CRITICAL | File inclusion, PHP filter | | 04 | Command Injection | CRITICAL | OS command execution | | 05 | SSRF | HIGH | Internal resource access | | 06 | Open Redirect | MEDIUM | Unvalidated redirects | | 07 | Directory Traversal | HIGH | Path traversal | | 08 | SSTI | CRITICAL | Template injection | | 09 | XXE | CRITICAL | XML external entities | | 10 | IDOR | HIGH | Access control bypass | | 11 | CSRF | MEDIUM | Missing tokens | | 12 | JWT Attacks | HIGH | 'none' alg, weak keys | | 13 | GraphQL | HIGH | Introspection, injection | | 14 | API Abuse | HIGH | Rate limiting, auth bypass | | 15 | File Upload | HIGH | Unrestricted upload | | 16 | Backup Files | HIGH | .env, configs, credentials | | 17 | .git Exposure | CRITICAL | Source code leak | | 18 | CORS | MEDIUM | Wildcard/reflective | | 19 | Race Condition | MEDIUM | TOCTOU, concurrency | | 20 | NoSQL Injection | HIGH | MongoDB $ne, $gt, $regex | ## Ollama Decision Engine Uses **granite4.1:8b** (change with `OLLAMA_MODEL` env var) to analyze recon data and pick the best vectors. The model sees: - What tech stack the target runs - What endpoints exist - What forms/inputs are present - What attack surfaces are visible Then it picks 5-10 vectors that have the highest probability of success. ## Reports Output in `reports/`: - `target.com_2026-06-18.md` — Full markdown report - `target.com_2026-06-18.html` — Styled HTML report with severity badges ## Requirements - `curl`, `jq`, `sqlmap` (recommended) - Ollama running (default `http://10.30.20.110:11434`) - Change with: `OLLAMA_HOST=http://localhost:11434 ./analyzer https://x.com` ## Deploy to Kali ```bash scp -r ~/the-analyzer root@10.30.20.177:/opt/ ``` ## Legal This tool is for **authorized bug bounty testing only**. Only use against targets you have explicit permission to test.