Skip to main content
Developer Toolv1.0.0

ZeroPort

A Windows app that answers one question fast: what is using this port? It lists every listening TCP port, reads the command line to name the real service — FastAPI, Next.js, Postgres — and stops it safely, revalidating process identity before any kill.

The ZeroPort window listing 22 listening TCP ports on Windows — each with its owning process, PID and inferred service, protected system ports locked and developer processes offering a stop action

How it works

  • Replaces the netstat → tasklist → guesswork loop with one window: every listening TCP port, the process that owns it, and a single action to release it.
  • Names the real service from the process command line rather than the port number, so "python -m uvicorn main:app" reads as FastAPI / Uvicorn instead of python.exe. The port is consulted last, so a bare python.exe on 5432 never becomes PostgreSQL.
  • Termination revalidates PID, start time and executable name immediately before acting, so a recycled PID can never be killed by mistake — and a process it cannot verify is marked un-stoppable rather than offered a stop button.
  • Sockets bound to both IPv4 and IPv6 collapse into one row per port and process: a real scan drops from 28 near-identical rows to 18, with the individual endpoints kept in the details view.
  • 135 tests run against the real operating system instead of mocks — binding live sockets, spawning and terminating real child processes, and driving the whole GUI end to end headlessly in CI.
  • Two runtime dependencies and zero network calls: no telemetry, no update check, no account. Shipping PySide6-Essentials instead of the full Qt meta-package kept ~140 MB of unused QML, WebEngine and 3D modules out of the binary.
  • Tagging a commit ships it — GitHub Actions runs the suite on a Windows runner, then publishes a single 34.6 MB executable with its SHA-256 only if every test passes.

Stack

Python 3.11PySide6 / QtpsutilPyInstallerpytestGitHub Actions

Role — Creator

Building something like this?

I build AI-powered products that ship to production — RAG pipelines, agentic workflows and the full-stack systems around them.

Get in touch