Generate intelligent context files that help AI assistants truly understand your project. Not just file dumps — structured analysis with prioritization.
curl -sSL https://livectx.com/install.sh | bash && livectx generate$ livectx generateTwo files that transform how AI understands your project
# SYSTEM_PROMPT.md
You are an AI assistant specialized in helping developers work with the **myproject** codebase.
## Project Overview
A Next.js e-commerce platform with real-time inventory management...
## Tech Stack
- Framework: Next.js 15 (App Router)
- Database: PostgreSQL with Prisma
- Styling: Tailwind CSS
## Key Files to Understand
1. `app/api/products/route.ts` - Product API endpoints
2. `lib/inventory.ts` - Real-time inventory sync logic
...See real output: CONTEXT.md and SYSTEM_PROMPT.md from the livectx repo itself
Simple commands, powerful output
Scanning /home/user/myproject... Found 70 files (234.8 KB) Skipped 5 files (ignored, binary, or too large) Languages: go (28), typescript (12), yaml (7), json (4), markdown (3) Analyzing codebase with AI... ✓ Project summary generated ✓ Key files identified (10 files) ✓ Architecture patterns detected ✓ System prompt generated Writing output files... ✓ CONTEXT.md (156 KB) ✓ SYSTEM_PROMPT.md (8 KB) Estimated tokens: ~60,109 Completed in 12.3s
Detecting AI coding tools... ✓ Claude Code ✓ Cursor ✓ GitHub Copilot Integrating with AI coding tools... ✓ Updated CLAUDE.md ✓ Updated .cursorrules ✓ Updated .github/copilot-instructions.md Done! Your AI tools will now reference the livectx context files.
Validating livectx configuration... Configuration: ✓ Configuration loaded successfully ✓ Config values valid API Key: ✓ ANTHROPIC_API_KEY configured Working Directory: ✓ /home/user/myproject ✓ .gitignore found ✓ .livectx.yaml found ✓ All checks passed!
Built for developers who want AI to truly understand their code
Uses LLM to analyze your codebase and identify key files, architecture patterns, and coding conventions.
Orders files by importance so AI assistants understand your project structure immediately.
Creates tailored AI assistant instructions specific to your codebase.
GitHub Action keeps your context files automatically updated on every push.
Your code stays between you and your chosen LLM provider. No data stored.
Scans and generates context in seconds, not minutes.
Most tools pack your files into a single document. livectx goes further — it uses AI to understand your codebase and generate context that actually helps AI assistants work effectively.
Entry points, core logic, and important configurations automatically detected
Custom AI instructions based on your project's patterns and conventions
Files prioritized by relevance, not alphabetically or by size
Tech stack, patterns, and coding conventions automatically documented
Give any AI assistant deep understanding of your codebase in one command.
Auto-generated context files that stay in sync with your codebase.
AI identifies what matters most so assistants focus on the right files.
1. Install livectx
$ curl -sSL https://livectx.com/install.sh | bash2. Generate context (prompts for API key on first run)
$ cd your-project && livectx generateGet your free API key at console.anthropic.com or openrouter.ai
Only to your chosen LLM provider (Anthropic or OpenRouter). livectx doesn't store, log, or transmit your code anywhere else.
~$0.01-0.05 per generation depending on codebase size. Caching automatically skips regeneration if nothing changed.
Any tool that reads markdown files — Claude Code, Cursor, GitHub Copilot, Windsurf, Aider, or just paste into ChatGPT. livectx integrate auto-configures popular tools.
Start generating intelligent context for your codebase today.