inkdust2021

inkdust2021 / VibeGuard

Public

VibeGuard, prevents you from leaking your privacy while vibecoding, letting you vibecode with peace of mind.

ai anthropic chatgpt claude cli
71
6
100% credibility
Found Feb 28, 2026 at 54 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Go
AI Summary

VibeGuard is a privacy protector that hides sensitive information like names or emails in your messages to AI coding tools and restores them in the responses.

How It Works

1
👀 Discover VibeGuard

You hear about a simple tool that keeps your personal info safe while chatting with AI coding helpers.

2
📦 Get it set up

Run the friendly installer script – it handles everything and asks what you need along the way.

3
⭐ Create your safety rules

Open the web dashboard on your computer, set a private password, and add words or patterns to hide like emails or names.

4
🔗 Link your AI tools

Launch your coding AI through the tool or point your app to the easy connection spot.

5
💬 Chat safely

Type away in your AI helper – it hides secrets before sending and brings them back in replies.

6
📊 Check the protection log

Peek at the dashboard to see what got hidden and confirm everything stayed private.

✅ Vibecode securely

Now you create code with AI without worrying about sensitive details leaking out.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 54 to 71 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is VibeGuard?

VibeGuard is a Go-built MITM HTTPS proxy that prevents leaking privacy while vibecoding by redacting sensitive keywords from requests to AI APIs like OpenAI or Anthropic, then restoring them in JSON or SSE responses. Set your IDE proxy to http://127.0.0.1:28657 or launch tools via CLI like `vibeguard claude` for process-only protection, letting you vibecode with peace of mind. Manage rules, audit hits, and stats through a local admin UI at /manager/.

Why is it gaining traction?

This legit VibeGuard stands out as a lightweight vanguard against accidental data leaks in AI prompts, with hot-reload rules, Docker isolation to keep CA keys off-host, and per-request audits showing exact matches without exposing originals. Unlike broad proxies, it's optimized for AI traffic—scans only text bodies up to 10MB, handles streaming SSE deltas across chunks, and offers process launchers that don't hijack your whole shell.

Who should use this?

Backend devs crafting proprietary algorithms in Cursor or Continue.dev, security-conscious teams prompting Claude with customer data, or solo hackers vibecoding personal projects without feeding secrets to cloud AIs. Ideal for anyone routing codex/gemini/opencode/qwen through a local proxy without setup headaches.

Verdict

Worth spinning up via Docker for privacy-focused AI coding—excellent docs, install scripts, and UI make evaluation painless despite 21 stars and 1.0% credibility score signaling early maturity. Test redaction thoroughly on real prompts before relying on it daily.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.