2 pointsby aragossa day ago1 comment
  • aragossa day ago
    Hi HN, I'm a QA engineer and the creator of Chaos Proxy. I built this because...

    I built this because debugging HTTP traffic with local proxies (like Charles or Proxyman) has become a bottleneck for me. Getting a GUI app running in CI is painful, and sharing a specific failing request with a colleague usually means sending screenshots or huge export files.

    Chaos Proxy is a managed MITM proxy running in the cloud. It solves the "it works on my machine" problem by giving you a shareable URL for your traffic logs.

    How it works:

    1. You get a dedicated proxy endpoint (e.g., `chaos-proxy.debuggo.app:8080`).

    2. Configure your device/simulator to point there and install the CA cert.

    3. Traffic flows through an isolated Docker container (running mitmproxy) where I apply chaos rules (latency, 500 errors) or simply inspect JSON.

    The Stack:

    - Core: Python & mitmproxy scripting API. Custom addons handle the chaos logic (bandwidth throttling, status code injection).

    - Orchestration: Next.js 16 (App Router) interacts with Docker Engine to spin up a fresh container for every project.

    - Sync: Redis acts as the bridge. Changing a delay rule in the UI applies it instantly to the running proxy process.

    - Storage: Supabase (Postgres) for metadata.

    Note on Sign-up: Since each session spins up a real, isolated Docker container, I require a sign-up to prevent abuse. However, no email verification or credit card is required — you can jump straight into the dashboard in 10 seconds.

    Privacy: Logs are ephemeral (TTL 24h) and bodies are not stored permanently. This is designed strictly for dev/staging, not production traffic.

    Happy to answer questions about the orchestration logic or how I handle the mitmproxy addons!