Strata tops the chart, HowToLiveBetter slips to two, and a passive radio observer debuts at four

Fastest-growing this period

Counts come from GitHub's official stargazers/history endpoint: the stars gained on each day of the window. Project briefs and the reading notes below are editorial additions.

1. 1

   Niko1221/Strata : A 125B-parameter model running on one gaming GPU, faster than you can read

   C++

   A one-click installer that runs Qwen3.8-Flash-Next, a 125-billion-parameter model that normally needs a server, on a single NVIDIA card with 12 to 24 GB of VRAM and 64 GB of RAM. It serves an OpenAI-compatible API on localhost, reports 60 to 95 tokens per second in its own measurements, accepts image input, and calibrates the engine to your PC.

   +1,174
2. 2

   eternity4719/HowToLiveBetter : Life advice that logs what you spend, what you gain and how solid the evidence is

   HTML

   A Chinese guide of practical suggestions on health, money, scams and legal traps, from chronic disease and first aid to rent disputes and workplace compensation. Each entry records what you spend, what you get back and how strong the evidence is, citing only journal papers and official documents, with an online search page for quick lookups.

   +947
3. 3

   DietrichGebert/ponytail : Stop your AI turning a small fix into a new framework

   JavaScript

   A set of instructions that asks AI coding agents to reuse existing code and question whether a new abstraction is needed. Its rules also explicitly retain validation and error handling, useful when your agent keeps making a small task unnecessarily complicated.

   +910
4. 4

   OffGridPete/Fieldwatch : A receive-only Wi-Fi and Bluetooth observer that keeps everything on your phone

   Kotlin

   A receive-only observer that lists every Wi-Fi access point and Bluetooth LE advertisement your Android phone picks up. No account, no backend and no dongle: scans, logs and reports stay on the handset, and a signature library names known trackers, cameras and other radio sources.

   +817
5. 5

   NVIDIA/OpenShell : Run autonomous agents inside kernel-enforced sandboxes

   Rust

   NVIDIA's runtime for fleets of autonomous agents. Declare in a policy what each agent may touch, and OpenShell enforces it in the kernel on every file access, system call and network connection, adding credentials only to requests bound for approved endpoints. Each policy change is formally verified before it takes effect, so newly granted risky access waits for human review.

   +776
6. 6

   ifixai-ai/iFixAi : An auditor that checks whether your AI agent actually does the job it was hired for

   Python

   An independent auditing tool for AI agents that grades whether an agent does the job it was hired for, not just whether it is fast or hard to hack. Guided setup picks the provider, judge and test suite, then about 60 inspections across five pillars return an A to F scorecard. Run it as a CLI wizard, as scripted flags in CI, or as a plugin that lets the agent audit itself.

   +766
7. 7

   Louis-CFM/coucou : A tiny friend in your notch that watches Claude Code and taps you when it needs you

   Swift

   A small open-source companion that lives in your Mac notch or at the top of a Windows screen and keeps an eye on every Claude Code session. Its animated character shows what each session reads, edits and runs, lets you approve or deny permission requests with one click, accepts dropped files, opens a chat with Claude, and reacts the moment a run finishes.

   +678
8. 8

   mattpocock/skills : Make your AI ask the right questions before it writes code

   Shell

   Give your coding assistant a skill that asks detailed questions before implementation. Use it to clarify vague requirements before the agent builds the wrong thing.

   +612
9. 9

   walkinglabs/learn-harness-engineering : A hands-on course on the machinery that keeps AI coding agents reliable

   TypeScript

   A project-based course of 14 lectures and 8 projects on the environment, state, verification and feedback machinery that makes AI coding agents work reliably, framed as five subsystems: instructions, tools, environment, state and feedback. It also reverse-engineers the real harness designs behind Claude Code, Codex, Pi and DeepSeek, in 15 languages.

   +610
10. 10

    debpalash/VoiceStudio : Try voice design, transcription and dubbing in one workspace

    Python

    A local audio workspace for voice design, transcription, dubbing and longer narrated projects. Useful for trying several speech workflows with your own material in one application, with engine choice and hardware determining the supported languages and output quality.

    +572

Ranking window: 2026-10-01 UTC, full day.