A reputation index from mitchellh's Vouch trust files
I was inspired by mitchellh's Vouch project, an explicit trust system where maintainers vouch for contributors before they can interact with a repo. Ghostty uses it to filter out AI slop PRs. Because Vouch exposes the vouch list as a plain text file (VOUCHED.td), I realized I could aggregate them across GitHub and build a reputation index. A crawler finds every VOUCHED.td file, pulls the entries, and computes a weighted score per user. Vouches from high-star repos count more than vouches from zero-star repos. Next step is to wire up an API so that the vouch GH action can start to use this…
In plain words
A reputation index that aggregates vouches from GitHub repositories using the Vouch trust system. The tool crawls VOUCHED.td files across projects and calculates weighted reputation scores for contributors, with vouches from popular repositories counting more heavily than those from smaller projects. It's designed for maintainers who want to identify trusted contributors and reduce low-quality pull submissions. An API is planned to integrate with GitHub actions for automated contributor approval.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
I was inspired by mitchellh's Vouch project, an explicit trust system where maintainers vouch for contributors before they can interact with a repo. Ghostty uses it to filter out AI slop PRs. Because Vouch exposes the vouch list as a plain text file (VOUCHED.td), I realized I could aggregate them across GitHub and build a reputation index. A crawler finds every VOUCHED.td file, pulls the entries, and computes a weighted score per user. Vouches from high-star repos count more than vouches from zero-star repos. Next step is to wire up an API so that the vouch GH action can start to use this data to auto approve contributors.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
