Yoke — Inteligência de Domínios e OSINT
yokedotlol/yoke
150+ sinais de inteligência de domínio — segurança, performance e reputação
Ver no GitHubMCP server para análise profunda de domínios cobrindo DNS, WHOIS, SSL/TLS, headers de segurança, stack tecnológico, performance (Lighthouse), breaches, subnets e scoring contextual em 6 eixos. Integra dados de múltiplas fontes (CertSpotter, HIBP, SecurityTrails) sem exigir autenticação e oferece análise com IA.
Instalação
npx convez-ai yokedotlol/yoke1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Investigação de segurança e OSINT de domínios
- Auditoria de postura de segurança (headers, SSL, SPF/DKIM/DMARC)
- Análise competitiva de stack tecnológico e infraestrutura
- Detecção de breaches e risco de reputação
- Validação pré-contratação ou due diligence
Veredito

Ferramenta madura e bem-documentada para pentesting, inteligência de ameaças e análise de segurança. O scoring contextual por arquétipo de site é diferenciam, mas demanda interpretação experiente — ideal para equipes de segurança, não para uso automático superficial.
README
A tradução para o português deste README ainda está na fila.
🔗 Yoke
Free, open-source domain intelligence — DNS, WHOIS, SSL, security, tech stack, performance, breaches, AI analysis, and more. Web, API, CLI, and Chrome extension.
Try it → yoke.lol · API Docs · CLI · Chrome Extension · Status
What is Yoke?
Yoke pulls 150+ scoring signals for any domain and presents them in a clean tabbed interface with a contextual scoring system. Think dig + whois + nmap + curl + BuiltWith + SecurityTrails — in one tool, no account required.
curl -s https://yoke.lol/stripe.com | jq
Scoring
6-axis contextual scoring: Security (0.24), Speed (0.18), Foundations (0.18), Reputation (0.15), Discoverability (0.13), Email (0.12).
Each axis starts at 100 and takes deductions per signal severity — budget-based deductive scoring, not additive. Tiers: Excellent ≥90, Strong ≥78, Moderate ≥60, Weak ≥40, Critical <40. An outlier floor cap prevents strong axes from masking a single weak one.
Sites are auto-classified into 7 archetypes (commerce, content, application, corporate, infrastructure, institutional, general) to adjust finding severity contextually. Compare domains side-by-side →
What It Checks
| Area | Highlights |
|---|---|
| DNS | A/AAAA/MX/NS/TXT/CNAME/CAA/SOA, TTLs, provider detection |
| WHOIS / RDAP | 4-tier resolution (RDAP → IANA → WhoisFreaks → raw), registrar, dates, age |
| SSL/TLS | Grade, issuer, protocols, key exchange, OCSP, CT logs, SSL Labs deep link |
| Security Headers | CSP, HSTS, X-Frame-Options, Permissions-Policy, Referrer-Policy, cookie audit |
| Email Auth | SPF, DKIM, DMARC, BIMI, MTA-STS, TLS-RPT |
| Performance | Lighthouse (mobile-first 60/40 blend), Core Web Vitals, CrUX, cache analysis |
| Tech Stack | 256 fingerprints — CMS, frameworks, CDNs, analytics, 25+ cookie consent platforms |
| WordPress | Version, theme, 100+ plugins, page builder, hosting |
| Breaches | HIBP lookup with time-decay severity weighting |
| Subdomains | 130 curated prefixes + CT log discovery via CertSpotter |
| Accessibility | 9 WCAG quick checks (labels, alt text, contrast, headings, landmarks) |
| Network | Global availability, TCP timing, BGP routing, WAF detection (29 providers) |
| Company | Wikidata + Brandfetch + Crunchbase enrichment, stock ticker |
| AI Analysis | Score Waterfall (deterministic) + Cross-Signal Insights (LLM, DeepSeek V3) |
Full signal registry: GET https://yoke.lol/api/scoring
API
No auth required. Content-negotiated — JSON for programmatic clients, HTML for browsers.
curl -s https://yoke.lol/stripe.com | jq # Full analysis
curl -s https://yoke.lol/stripe.com | jq '.ssl' # Specific section
curl -s https://yoke.lol/stripe.com | jq '.domain_score' # Scoring breakdown
Rate limits: 20 analyses/hr per IP (cached results don't count). Full API docs →
| Method | Path | Description |
|---|---|---|
GET | /:domain | Full analysis (JSON) |
POST | /api/analyze | Analysis (JSON or SSE via Accept: text/event-stream) |
POST | /api/compare | Side-by-side domain comparison |
POST | /api/subdomains | CT log subdomain discovery |
POST | /api/subdomain-scan | Active subdomain enumeration |
POST | /api/ai-analysis | AI deep analysis (Cross-Signal Insights) |
GET | /api/scoring | Scoring methodology + full signal registry |
GET | /api/health | Service health |
GET | /api/docs | API documentation (JSON) |
Additional endpoints: /api/company, /api/news, /api/social, /api/suggestions, /api/reverse-ip, /api/availability. See API docs for details.
CLI
Fast Go-based CLI for terminal domain analysis.
# Install
brew install yokedotlol/tap/yoke # Homebrew
curl -sSL https://yoke.lol/install.sh | bash # Shell script
# Use
yoke stripe.com # Full analysis card
yoke stripe.com --json | jq # Raw JSON
yoke score stripe.com # Quick score
yoke compare github.com gitlab.com # Side-by-side
yoke ai stripe.com # AI analysis
# Point at your own instance
yoke config --set-base-url https://your-instance.com
The CLI auto-checks for updates via X-Yoke-Min-Client header.
Chrome Extension
Click the Yoke icon on any site to open a side panel with full analysis. To load from source: chrome://extensions/ → Developer mode → Load unpacked → select extension/.
MCP Server
For AI coding agents. 3 tools: yoke_analyze, yoke_score_summary, yoke_compare.
npm install -g @yokedotlol/mcp-server
See mcp/README.md for configuration.
Badges
Embed a live domain score badge in your README, docs, or website. Badges are auto-refreshed — scores stay within 24 hours of the latest analysis.
Direct SVG (standalone, no external dependencies):

Shields.io endpoint (use shields.io's rendering):

Options:
| Parameter | Example | Description |
|---|---|---|
axis | ?axis=security | Show a specific axis score |
label | ?label=My+Site | Override badge label |
style | ?style=flat-square | Sharp corners |
Self-hosters: badges work out of the box. See Self-Hosting docs for pre-warm cron setup.
Architecture
┌─────────────────────────────────────────────────┐
│ yoke.lol │
│ │
│ ┌───────────┐ ┌────────────────────────┐ │
│ │ React SPA │──────▶│ Cloudflare Worker │ │
│ │ (Tailwind,│ │ (zero-dep router) │ │
│ │ RQ, │ │ │ │
│ │ Leaflet) │ │ ┌──────┐ ┌────────┐ │ │
│ └───────────┘ │ │ D1 │ │ KV │ │ │
│ │ │stats │ │ cache │ │ │
│ ┌───────────┐ │ └──────┘ └────────┘ │ │
│ │Chrome Ext.│─iframe─┘ │ │
│ └───────────┘ │ │
└──────────────────────────┼─────────────────────┘
│
┌────────────────┘
│ ┌─────────────────────┐
│ │ Fly.io Proxy (Go) │
│ │ HTTP probes, GeoIP │
│ └─────────────────────┘
│
└── 20+ external APIs (DNS, RDAP, SSL Labs,
HIBP, Shodan, PageSpeed, Wikidata, etc.)
| Layer | Tech | Size |
|---|---|---|
| Frontend | React 19, Tailwind v4, React Query, Leaflet | ~213KB initial JS |
| Backend | Cloudflare Worker (TypeScript, zero deps) | ~214KB bundled |
| Database | Cloudflare D1 (stats) + KV (cache, rate limits) | — |
| Proxy | Go on Fly.io (HTTP probes, MaxMind GeoIP) | — |
| Build | Bun + Vite, Node.js 22+ for Wrangler | — |
Self-Hosting
Yoke is designed to be self-hosted. Three deployment options:
| Option | Best For | Cost |
|---|---|---|
| Cloudflare Workers | Managed hosting, global edge | ~$5–11/mo |
| Docker Compose | VPS self-hosting, one command | ~$5–18/mo |
| Bare Metal | Full control, no containers | ~$5–18/mo |
Quick start (Cloudflare Workers):
git clone https://github.com/yokedotlol/yoke.git && cd yoke
cd client && bun install && cd ..
cd worker && bun install && cd ..
npx wrangler d1 create yoke-stats
npx wrangler kv namespace create REFERENCE_DATA
cp worker/wrangler.toml.example worker/wrangler.toml # edit with your IDs
bash deploy.sh --cf
Quick start (Docker Compose):
git clone https://github.com/yokedotlol/yoke.git && cd yoke
cp .env.example .env # edit with your domain + secrets
docker compose up -d --build
Contributing
Contributions welcome — the easiest entry point is adding a new analysis check (one file, standard interface). See CONTRIBUTING.md.
AI Contributors
Start with CLAUDE.md for technical context. The .ai/ directory has structured project context (architecture, invariants, decisions, gotchas). Run bash .ai/staleness-check.sh to verify docs match the codebase.
The .ai/reviews/ directory contains a 20-panel expert review system you can run with any AI agent harness — see .ai/reviews/README.md.