RedForge is an autonomous security orchestration platform that runs real HTTP probes against target URLs — not simulations. It coordinates 11 detection modules in parallel, correlates findings into multi-stage attack chains, and generates ready-to-paste remediation code for every vulnerability it finds.
⚠️ Ethical use only. Only scan systems you own or have explicit written permission to test.
The engine (scanner/index.ts) operates in 6 sequential phases, with the core detection running fully in parallel:
Phase 1 ── Target fingerprinting (tech stack, HTTP status, headers)
Phase 2 ── Parallel module execution (11 modules fire simultaneously)
Phase 3 ── Active probing, SQLi injection, rate-limit tests [ACTIVE mode]
Phase 4 ── AI deep analysis via NVIDIA NIM (Nemotron-70b)
Phase 5 ── Attack chain correlation (7 multi-stage exploit patterns)
Phase 6 ── Deduplication, risk scoring, enrichment, scan diff
All 11 modules run via Promise.all — no sequential bottleneck:
┌─────────────────────────────────────────────────────────┐
│ TLS/Cookies │ Headers │ Info Disclosure │ Auth │
│ Supply Chain │ XSS │ SSRF/Redirect │ DNS │
│ API Security │ WordPress │ ← 11 modules │ │
└─────────────────────────────────────────────────────────┘
↓ all findings merged & deduplicated
Attack Chain Correlation Engine
↓
Enrichment Pass (remediation + CVE + compliance + diff)
RedForge doesn't just list individual findings — it identifies multi-stage exploit chains by correlating finding tags:
| Chain | Tags Required | Severity |
|---|---|---|
| XSS + Session Hijack | xss + httponly |
CRITICAL |
| SSRF + Cloud Metadata | ssrf + aws/gcp |
CRITICAL |
| WordPress + RCE | xmlrpc + wp |
HIGH |
| Credential Stuffing | rate-limit + auth |
HIGH |
| Supply Chain Injection | sri + xss |
HIGH |
| # | Module | What it detects |
|---|---|---|
| 1 | TLS/Cookies | HSTS, secure/HttpOnly flags, weak cipher suites |
| 2 | Headers | CSP, X-Frame, Referrer-Policy, Permissions-Policy |
| 3 | Info Disclosure | Server version banners, stack traces, debug logs |
| 4 | Auth Security | Login panel exposure, default credentials, JWT issues |
| 5 | Supply Chain | Missing SRI on CDN scripts, dependency confusion |
| 6 | XSS Detection | Reflected, DOM-based, stored XSS patterns |
| 7 | SSRF/Redirect | Open redirects, meta-refresh, SSRF parameter probing |
| 8 | DNS Security | SPF, DMARC, DKIM, CAA record validation |
| 9 | API Security | GraphQL introspection, WAF fingerprinting, directory traversal |
| 10 | WordPress | Version disclosure, XML-RPC, user enum, plugin discovery |
| 11 | AI Analysis | NVIDIA Nemotron-70b deep pattern analysis |
After detection, every finding is automatically enriched:
- Auto-Remediation — 20 fix templates (nginx / apache / express / DNS) attached as ready-to-paste code
- CVE Enrichment — NVD API v2 lookup for detected software versions with CVSS scores
- Compliance Mapping — GDPR articles, PCI DSS, ISO 27001, NIST SP800-53, DPDP Act 2023 (India)
- Scan Diffing — new/resolved/regressed findings vs previous scan, with real fix rate %
| Layer | Tech |
|---|---|
| Frontend | React 18 + Vite + Tailwind CSS v4 |
| Backend | Express 5 + Node.js |
| Database | PostgreSQL + Drizzle ORM |
| AI | NVIDIA NIM — nvidia/llama-3.1-nemotron-70b-instruct |
| Deployment | Vercel (static CDN + serverless function) |
| Monorepo | pnpm workspaces |
# Install
pnpm install
# Push DB schema
pnpm --filter @workspace/db run push
# Seed demo data
pnpm --filter @workspace/scripts run seed
# Run (both servers)
pnpm run devRequired env vars:
DATABASE_URL=postgresql://...
SESSION_SECRET=your-random-secret
NVIDIA_NIM_API_KEY=nvapi-... # enables AI analysis (Nemotron-70b)
APP_URL=https://redforgex.vercel.app| Mode | What runs |
|---|---|
| PASSIVE | All 11 parallel modules + correlation engine |
| ACTIVE | + SQLi probes, rate-limit tests, business logic checks |
MIT — for educational and portfolio purposes. Scan only systems you own or have permission to test.
