nowfound

AI · August 28, 2025

IC

I cut my Claude API bill by 66% with Git-based context

Hey HN, I built ShadowGit a while back to automatically commit code every minute to a hidden git repo (.shadowgit.git). Original goal was to easily rollback when AI tools break things. But I discovered something interesting: this minute-by-minute history is perfect context for AI assistants. So I built an MCP server that lets Claude/Cursor query this history using native git commands. The results surprised me: Before: Claude would read my entire codebase repeatedly, burning 15,000+ tokens to debug issues. After: Claude runs `git log --grep="drag"` finds when drag-and-drop worked,…

What it does

In the maker’s words, at launch

Hey HN, I built ShadowGit a while back to automatically commit code every minute to a hidden git repo (.shadowgit.git). Original goal was to easily rollback when AI tools break things. But I discovered something interesting: this minute-by-minute history is perfect context for AI assistants. So I built an MCP server that lets Claude/Cursor query this history using native git commands. The results surprised me: Before: Claude would read my entire codebase repeatedly, burning 15,000+ tokens to debug issues. After: Claude runs `git log --grep="drag"` finds when drag-and-drop worked, applies that fix. 5,000 tokens. How it works technically: 1. ShadowGit watches your project and commits every save to .shadowgit.git (parallel to your main repo) 2. MCP server exposes git operations to AI: - `git diff HEAD~10 -- component.tsx` to see recent changes - `git log -S "functionName"` to find when code was added - `git show COMMIT:file.js` to retrieve working versions 3. AI naturally understands git, so instead of dumping context, it surgically queries what it needs It's free while I am testing and improving it. I am looking for feedback on the project and maybe some integration ideas beyond MCP. Thanks! Alessandro

Does the same job

all alternatives →
  • Context GatewayMar 2026 · ▲220

    Make Claude Code faster and cheaper without losing context

  • claude-devtoolsFeb 2026 · ▲144

    See everything Claude Code hides from your terminal

  • CanopyJun 2026 · ▲77

    Parallel, sandboxed Claude Code sessions on native macOS

  • CC
    Claudit – Claude Code Conversations as Git Notes, AutomaticallyFeb 2026 · github.com · ▲7

    Uses agent and Git Hooks to automatically create Git Notes on commit, containing the agent conversation that led to that commit. Works if either you or the agent commit. It's basically the same thing as entire.io just announced that they got $60m investment for. Except I got Claude Code to write it last week, in my spare time, without really paying attention. I certainly didn't read or write any of the code, except for one rubbish joke in the README. I've got a Claude Code instance working on Gemini CLI support and OpenCode support currently.

  • MaestroJan 2026 · ▲20

    Bloomberg Terminal for CLI agents

  • CC
    Ccgs – Collaborative Claude Code sessions, stored in Git branchesJun 2026 · github.com · ▲6

    My team uses Claude Code daily, and the sessions have become some of the most useful artifacts we produce. But they're trapped in ~/.claude/projects/ on whichever laptop they happened on. There's no good way to hand a colleague "the session where I untangled the migration" so they can claude --resume it and keep going from where I left off. Enter ccgs: Share Claude Code sessions through an orphan branch (@ccgs/) in your existing repo's remote - Session files carry the author's absolute paths. On pull, ccgs rewrites the working dir back to your path so resume actually…

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

  • Astute585

    Automate your B2B brand going viral, with new media creators

    AI · 18d ago · company-app.joinastute.com

  • Grok Bot547

    AI teammates that you can give real work to

    AI · 25d ago · x.ai

  • 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

  • Make your software self-driving

    AI · 30d ago · coldtea.ai

  • Soloop472

    Approval-first Agent OS for solo founders

    AI · 30d ago · soloop.io

Launched alongside, August 2025

the whole month →
  • Trace1,556

    Workflow Automations for the Human 👾 AI Workforce

    AI · 2025 · trace.so

  • IS

    I built the world's most impractical 1000-pixel display and anyone in the world can draw on it. It draws a single pixel at a time and takes 30-60 minutes to complete a single image. Anyone can participate in the project by voting for the next image to be drawn, and submitting images. https://kilopx.com/

    Work · 2025 · benholmen.com

  • Job boards are dead. Your network is alive

    AI · 2025 · contra.com

  • KT

    Kitten TTS is an open-source series of tiny and expressive text-to-speech models for on-device applications. We are excited to launch a preview of our smallest model, which is less than 25 MB. This model has 15M parameters. This release supports English text-to-speech applications in eight voices: four male and four female. The model is quantized to int8 + fp16, and it uses onnx for runtime. The model is designed to run literally anywhere eg. raspberry pi, low-end smartphones, wearables, browsers etc. No GPU required! We're releasing this to give early users a sense of the latency and voices…

    Dev tools · 2025 · github.com

  • IW

    I was wondering how I can arrange objects along a spherical helix path, and read some articles on it. I ended up learning about parametric equations again, and make this visualization to document what I learned: https://visualrambling.space/moving-objects-in-3d/ feel free to visit and let me know what you think!

    Life & fun · 2025 · visualrambling.space

  • TC

    For HTML Day 2025 [1], I made a web service that displays the current sky at your approximate location as a CSS gradient. Colours are simulated on-demand using atmospheric absorption and scattering coefficients. Updates every minute, without the use of client-side JavaScript. Source code and additional information is available on GitHub: https://github.com/dnlzro/horizon [1] https://html.energy/html-day/2025/index.html

    Dev tools · 2025 · sky.dlazaro.ca