← back to the directory
Application Coding

Micro-agent

1AI 2writes 3code 4passing 5your 6tests

the six Ws · specification

W1 Who

Maintained by Builder.io as an open-source project.

W2 What

A CLI agent that iteratively writes and fixes code until it passes the tests you provide.

W3 Where

Runs locally as a Node CLI in your project directory.

W4 When

Reach for it when a task is expressible as tests and you want a passing implementation.

W5 Why

The test-driven loop constrains the AI, producing verifiable, focused code.

W6 With

An LLM API key (OpenAI or Anthropic) and Node.js.

W7 Watch

Vendor-backed open-source; runs your test commands and edits files locally in a bounded loop. MIT licensed, small but stable.

Test-drivenCodegenCLITypeScript

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.