I've been working on llmcc, a tool that generates architecture view at multiple "zoom levels" for codebases. The idea: coding agents (and humans) need to understand systems from different distances—sometimes you need the high-level crate/module boundaries, sometimes the low-level symbol relationships.
It uses PageRank to surface the most "important" nodes, so you can get a focused view of a 600K LoC codebase in seconds.
Currently Rust-only, but the architecture is designed to support multiple languages.
Built this because I kept seeing coding agents get lost in large codebases—they'd either get too zoomed in (missing context) or try to process too much. Having a tree-like view where you can walk up/down seemed like the right primitive.
Repo: https://github.com/allenanswerzq/llmcc, checked out for svgs or benchmark if interested.
Looking for feedbacks:
- Do you guys feel this could be useful and continue spend time to for it?
- What languages would you want next?
- Any kinds of ideas regarding this project for coding agents?
Happy to answer questions!