Built for developers
who value clarity.

The problem we're solving

When you inherit a codebase — or revisit your own project after a few months — the first hours are always the same. You clone it, start reading files, chase imports, wonder why certain decisions were made, and slowly build a mental model that should have been documented from the start.

We think that mental model should be instantly available, not something you reconstruct from scratch every time. ContextGraph builds it for you: a living, interactive map of how your code is actually organized.

What we actually do

Paste a GitHub URL. We fetch the repository, parse the file tree, resolve imports and exports, detect entry points, and lay it all out as a navigable architecture graph. Click any node and an AI-generated explanation tells you what that module does and why it matters.

• Static analysis across TypeScript, JavaScript, Python, Rust, Go, and more
• Dependency graph with complexity scoring per module
• Entry point detection — where does the code actually start?
• AI explanations grounded in the actual code, not hallucinations
• Side-by-side repository comparison to spot shared patterns

Where we are

ContextGraph is in public beta. We've analyzed over 10,000 repositories and mapped more than 100,000 individual files. We're working on private repo support, team workspaces, and a VS Code extension. Follow @contextgraphart for updates.

Values

• Speed over perfection. A fast, useful answer beats a slow perfect one.
• Honesty. We show confidence scores and say "unclear" when we don't know.
• Developer-first. We build tools we'd want to use ourselves, every day.
• Privacy by default. We don't store your source code. We analyze and discard.