RAG-powered codebase search for AI coding assistants

Stop guessing which files
to paste into your AI.

Context Compiler indexes your GitHub repositories into a semantic vector store. Search with natural language, ask questions about your code, and assemble precise context snippets for Claude, Cursor, or any AI assistant.

View pricing →

Everything you need to give AI the right context

Natural Language Search

Find any function, pattern, or concept in your codebase with plain English. No regex, no grep — just ask.

Repository Q&A

Ask questions about your codebase and get AI-generated answers grounded in actual code, with source citations.

Context Cart

Collect relevant code snippets into a Context Cart and export the whole set as a token-optimized prompt for any AI assistant.

Incremental Re-indexing

Only changed files are re-embedded on re-scan. Unchanged chunks reuse cached vectors — fast updates without full re-indexing.

GitHub App Sync

Webhook-driven repository inventory with automatic reconciliation. No polling, no manual imports.

Team Workspaces

Share saved context carts and answer sessions across your team. Role-based access for owners, admins, and members.

How it works

01

Connect your repositories

Sign in with GitHub and install the Context Compiler App. Your repository inventory syncs automatically via webhooks.

02

Scan and index

Trigger a scan and the AI microservice clones, chunks, and embeds your codebase into a pgvector HNSW index.

03

Search, ask, and build context

Use natural language search or repo Q&A to find what you need, then collect it in the Context Cart for your AI coding session.

Start for free. No credit card required.

Index one repository and run 100 searches per month at no cost. Upgrade when you need more.

See all plans