MiniApp Runner in Browser

Runs WeChat mini-programs directly in vanilla Chrome and headless browsers for AI agent control and CI testing.

Problem

Developers testing mini-programs find the official WeChat devtools too heavy: several GB of memory, frequent re-login via QR code, and painful automation in Linux CI.

Solution

A CLI that runs mini-programs in regular Chrome/headless browser using the official base library, closely matching real-device behavior. Expose DOM, console and network requests so Claude Code, Codex and other agents can operate pages directly, and support Playwright tests.

Users

Mini-program developers, AI coding agent users, teams needing mini-program automation testing.

MVP

CLI launches local Chrome to load a mini-program bundle; hot reload; expose page DOM and console logs; headless mode and direct URL to any page; output a connectable endpoint for Playwright.

Revenue

Open-source core CLI for adoption; paid cloud device farm, parallel test acceleration and team collaboration dashboard, or per-test-minute billing.

Why now

AI coding agents operating browsers is a growing workflow, while mini-program devtools remain closed and heavy. This gap is amplified by agent-driven automation demand, and an indie developer can move fast.