I'm building TunnelScope, an MIT-licensed Python CLI for evidence-based MTU and packet-size experiments. Linux senders and Linux or Windows receivers retain observations in JSON, with offline HTML and bounded Markdown for review.
Update — 8 November 2026: v0.9.3 alpha is on PyPI. Several starter tasks have now been completed by contributors. The 30-second recorded evidence replay shows the sharing workflow.
TunnelScope is still an alpha tool: a timeout is inconclusive, and an observed payload interval is not an independently measured end-to-end PMTU or proof of a root cause. Its UDP protocol and receiver logs are unauthenticated, so experiments belong on permitted controlled paths.
Recent completed contributions
- Python 3.13 delivery coverage: all six Linux/Windows matrix jobs pass; issue #33 is completed.
- Reject a joined report used as raw sender input: preserves legacy inputs and checks that errors create no output files; issue #62 is completed.
- Retain the IPv6 receiver's local scope: adds local metadata without treating it as a cross-host identity; issue #64 is completed.
Check the live open task list for current work. A small reproducible bug report or focused documentation improvement is welcome too. Please comment before starting a listed task, and discuss larger changes first. The original Python 3.10 CI, receiver quick-start and PowerShell documentation tasks from this post were completed earlier.
Field reports are welcome too
Controlled records cover WireGuard, OpenVPN, IPv6 feedback, Windows/WSL2 host/guest paths and one Kubernetes/Calico setup. Those checks do not replace an independent operator VPN report or physical multi-host evidence. A normal working path is useful too. The field guide explains bounded collection, joining and sanitizing related reports together. Review copies before sharing; no remote access, credentials or private packet captures need to be shared.
Thanks to sunxiuguo for PRs #53, #56 and #66, and HarshRajSinghania for #61 and #67. Five PRs from two contributors have merged after maintainer review and delivery checks.
Start with the README and contribution guide. Reproducible reports, documentation feedback and focused PRs are welcome.
Top comments (0)