Your AI agent's test engineer. It writes brutal test suites that try to break your code.
How to pay
Subscribe
$9.99/month
Predictable monthly cost with included usage. Best for steady, high-volume traffic.
- Unlimited tools within plan limits
- One API key, billed once a month
- Cancel any time
Edgehound is the test engineer for AI-built code. It writes brutal test suites that try to break what your agent just created - then your agent runs them, fixes what fails, and reruns until the suite passes. Automatic from the moment it is installed.
HOW IT WORKS
- Install it in Claude Code, Cursor, or any MCP client. One URL.
- Your agent calls it automatically whenever it writes or changes code that handles input, money, dates, or state. No prompting needed.
- The suite runs. Failing tests are real bugs: plus-addressed emails, emoji in strings, timezone and month-end math, money rounding, double submits and races, injection strings, pagination edges.
- Your agent patches the code, reruns, and reports what it caught. You approve and ship.
WHAT IT COVERS 150+ real failure patterns across 14 categories, delivered as runnable suites for Jest, Vitest, node:test, pytest, Go, PHPUnit, RSpec, or cargo test. Every run comes with a coverage report showing which categories were exercised and where a failing test points at a real bug.
Built for people shipping AI-written code, where the happy path takes seconds and the ugly cases never get tested.