AI Badger — Local AI Coding Context Tool

Local-first codebase context extraction for any AI chat (Claude, ChatGPT, Grok, DeepSeek, etc.).

Get precise, token-efficient context on demand without uploading your entire repo or wasting tokens on irrelevant files.

No cloud • No API keys • No telemetry • Fully local

▶ Try Interactive Demo • Install

Map → Extract → Apply: Smart local context bridge that prepares focused codebase snippets for any LLM chat.

How it works

1. Map
Enter your goal. Badger builds a prompt.
↳ You copy it → paste into your AI chat

2. Extract
AI replies asking for specific files.
↳ You copy that → paste back into Badger

3. Apply
Badger fetches those files, builds a second prompt.
↳ You copy it → paste into AI → review before writing

✓ Fully local — nothing leaves your machine until you copy it
✓ You control every paste and every write

Perfect for Claude token saving, local LLM workflows, code reviews, design sessions, and debugging.

Why AI Badger?

• Universal compatibility — Works with any AI chat interface or local model
• Local-first codebase context tool — Complete privacy, no uploads
• Token efficient — Stop burning agent tokens on reviews, explanations, or brainstorming
• Precise & lightweight — Built in Go, fast, minimal overhead
• Specialized modes — review and design for common workflows

Install

Homebrew (Recommended)

brew install pvrlabs/tap/badger

Quick Curl Install

curl -fsSL https://raw.githubusercontent.com/PVRLabs/aibadger/main/install.sh | sh

See docs/install.md for Windows, source builds, and more.

Also available with an official VS Code companion.

Agent Skills

Badger includes the handoff and badger-review skills for continuing an AI coding session or requesting an independent review. See the Agent Skills guide for installation, usage, and details.

Quick Start

1. Run badger in your project root. Interactive sessions start in Design focus.
2. Type your goal (or leave the editor empty and press Enter to explore the project).
3. Copy Prompt 1 → paste into your AI chat.
4. When the AI asks for files, copy its response → paste back into Badger.
5. Copy Prompt 2 → paste back to the AI.
6. Paste the AI’s response into Badger → review and apply changes.

Specialized Modes

• badger code — explicitly start in Code focus
• badger review — Git changes and bounded supporting context for immediate findings
• badger design — explicitly start in Design focus with an empty editor

Full usage: docs/usage.md

Learn More

• Usage Examples & Walkthrough
• Browser Handoff Guide
• API Reference — Non-interactive commands for editors and scripts
• Agent Integrations — Compact repository orientation for coding agents
• Articles
• Protocol Reference
• Limitations & Supported Projects
• Privacy & Safety
• Contributing

Star if this local AI coding context tool solves a real pain for you ⭐

Built in San Diego by PVR Labs. 🌊

Website • X @kupolov