The Journey
I spent the last week building TestForge — a testing platform that analyzes code across 21 dimensions. Not 3 like Snyk. Not 5 like SonarQube. Twenty-one.
Here is everything that was built, tested, and deployed to production.
What TestForge Analyzes
Standard Dimensions
- Security scanning — SAST with OWASP Top 10 coverage mapping
- Unit test analysis — Coverage estimation, framework detection
- Load & performance — Rate limiting, caching, connection pooling
- Accessibility — WCAG compliance, alt text, form labels
- Contract testing — OpenAPI/Swagger detection, API versioning
- Visual regression — Style consistency
Strategic (Only TestForge Has These)
- 🤖 Agentic Scale Prediction — World first. Simulates thousands of AI agents hitting your API. Predicts failure patterns, bottlenecks at scale.
- 📊 DORA Metrics — Deployment frequency, lead time, MTTR, change failure rate
- 🎯 Vision & Goal Alignment — Observability, feature flags, analytics, CI/CD maturity
- 📦 Scope Coverage — Documented vs implemented features
- 🏗️ Stack Analysis — Architecture quality, TypeScript, ORM, caching
Deep Analysis
- N+1 Query Detection — DB queries inside loops
- Supply Chain Audit — CVE database cross-reference
- Dead Code Detection — Unused dependencies
- License Compliance — GPL/copyleft risk flagging
- Edge Cases + Property-Based + Chaos + Mutation + Predictive
Real Test Results (tinyhttp/malibu)
Security 4 findings ████░░░░░░░░░░░░░░░░
Unit Tests 67% █████████████░░░░░░░
Agentic Scale 24/100 ████░░░░░░░░░░░░░░░░ ⚠️ Critical
DORA 30/100 ██████░░░░░░░░░░░░░░ ⚠️ Manual deploys
OWASP Coverage 20% ████░░░░░░░░░░░░░░░░
Top findings: Auth gaps, no rate limiting, no observability, 34 undocumented endpoints.
Tech Stack
| Layer | Technology |
|---|---|
| Frontend | React 19 + TypeScript + Vite + Tailwind |
| Backend API | Vercel Serverless (18 endpoints) |
| MCP Server | Fastify + TypeScript on Fly.io |
| Database | Neon PostgreSQL + Drizzle ORM (8 tables) |
| Payments | Stripe Checkout + Webhooks |
| Auth | GitHub OAuth |
| npm | @whitenoisenpm/testforge-mcp v0.2.3 |
Enterprise Features
- Data Isolation: Per-user repos, tests, API keys (7 tasks)
- Security: Rate limiting, API key auth, CORS, security headers (6 tasks)
- Dashboard: Overview, repos, test history, API keys, billing, team (12 tabs)
- Reliability: Auto-scaling, health checks, DB backups (6 tasks)
- UX: Onboarding modal, empty states, error boundaries, toasts (6 tasks)
Try It
# Web Platform
open https://testforge.run
# MCP IDE Integration
npx @whitenoisenpm/testforge-mcp install
# CLI Score (CI/CD friendly)
npx @whitenoisenpm/testforge-mcp score https://github.com/user/repo
# Self-Hosted
git clone https://github.com/t4tarzan/testforge
cd testforge/mcp-server && flyctl launch --now
Pricing
| Plan | Price | Tests |
|---|---|---|
| Free | $0 | 5/mo |
| Pro | $29/mo | 100/mo |
| Enterprise | $199/mo | Unlimited |
Website: testforge.run | GitHub: t4tarzan/testforge
Built by t4tarzan — Ship code you trust. 🚀
Top comments (0)