Skip to content
2nd Place

Cache Cow

Cache Cow is a VS Code extension that solves the critical, post-generation problem of Vibe Debugging (Problem Statement Three).

Code Video Cerebral Valley

Core trick

A VS Code extension sends diffs and infrastructure plans to Gemini, returning cost, latency, and metric scores shown in a sidebar.

Tags

dev tools

Description

Cache Cow is a VS Code extension that solves the critical, post-generation problem of Vibe Debugging (Problem Statement Three). While "vibe coding" with AI is fast, it often introduces hidden costs, latency, and quality issues that only appear in production. Cache Cow acts as an immediate, proactive quality gate, using Gemini to analyze code changes and infrastructure plans before they are committed. It transforms dry, technical analysis into energetic, motivational "Vibe Checks" delivered through a custom VS Code sidebar, making code analysis fun and constructive.
vishalv971/cache-cow★ 10 forksJavaScriptMITCreated Nov 15, 2025 · last push Nov 15, 2025

README excerpt

# 🐄 Cache Cow

AI-powered code analysis for cost, latency, and metrics using Google Gemini.

## Features

- **Cost Impact Analysis**: Identifies code changes that could increase computational costs
- **Latency Analysis**: Finds potential performance bottlenecks
- **Code Metrics**: Evaluates code quality, complexity, maintainability, and security
- **AI Recommendations**: Get actionable suggestions powered by Gemini AI

## Setup

### 1. Install Dependencies

```bash
npm install
```

### 2. Set up Gemini API Key

Get your Gemini API key from [Google AI Studio](https://ai.google.dev/).

**Configure in VS Code Settings:**
1. Open VS Code Settings (`Cmd+,` or `Ctrl+,`)
2. Search for "Cache Cow"
3. Enter your Gemini API key in the `cacheCow.geminiApiKey` field
4. (Optional) Enter your Prometheus server URL in the `cacheCow.prometheusUrl` field (e.g., `http://localhost:9090`)

### 3. Run the Extension

1. Press `F5` to start the Extension Development Host
2. Look for the Cache Cow icon in the left Activity Bar
3. Click it to open the sidebar
4. Click "Run Analysis" to analyze your code

## How It Works

1. **Collects Code Data**: Reads your git diff and file contents (respects `.gitignore`)
2. **Fetches Prometheus Metrics** (Optional): If configured, fetches current production metrics like CPU usage, memory, request latency, error rates, etc.
3. **Sends to Gemini**: Analyzes the code using Gemini 2.0 Flash, including production metrics for context-aware analysis
4. **Displays Result

Team