One tool. Zero overhead.
No servers. No databases. No API keys. Just your code — packaged and ready for any AI.
Zero Infrastructure
No servers, no vector databases, no API keys. Run it once, get a file. That’s it. Zero cost, zero setup, zero maintenance.
$0
Infrastructure cost
Optimized for LLMs
Strips repetitive code, adds metadata, and generates token-optimized reports. Split huge codebases into chunks with --chunk-size.
~68%
Token reduction
Live Context with Watch Mode
OS-level file events update your report the moment you save. Your LLM always has the latest code.
Works with Any LLM
ChatGPT, Claude, Gemini, local models. One file, every AI.
Multi-Root Projects
Frontend, backend, docs — scan them all simultaneously. Multiple paths generate individual reports plus a unified combined.html dashboard.
HTML Dashboard
Interactive single-file dashboard with charts, virtual-scroll source viewer, and syntax highlighting. No server required.
Upload to ZagForge
Push your scan result to ZagForge with --upload. Authenticates via ZAGFORGE_API_KEY or ~/.zagforge/credentials.
Built with Zig
Extreme performance for massive codebases. Run zigzag bench for a per-phase timing breakdown.
What happens when you run ZigZag.
0.8s
3,763 files scanned
4 formats, one pass
Why not just copy-paste?
You could. But here's what you'd be missing.
Questions? Answers.
Everything you need to know before getting started.