iantbutler01

iantbutler01 / code_diver

Public

LLMs write a lot of code, but humans still need to understand it...for now. Code Diver provides 2 things a skill that teaches your agent how to build a connected graph of higher level semantically related code as it works and then ingests, builds and visualizes this graph in realtime so you can review massive changes more easily.

39
2
100% credibility
Found Mar 19, 2026 at 39 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Code Diver is a local visualization tool that maps code projects by blending human-written notes with automatic structure detection into an interactive graph for exploration and review.

How It Works

1
📖 Discover Code Diver

You hear about a friendly tool that turns confusing code folders into easy-to-explore maps for reviews and handoffs.

2
🛠️ Get it ready

Download and prepare the viewer on your computer with a few simple steps.

3
📁 Pick your project

Choose the folder with the code you want to understand.

4
🚀 See the magic map

Launch the tool and watch a colorful, interactive graph of files, parts, and connections build right before your eyes.

5
🧭 Wander and zoom

Click on groups, files, or notes to dive deeper and uncover hidden relationships.

6
🔍 Spot changes

Hover over items to reveal what's new, updated, or removed in your project.

🎉 Project makes sense

You now have a clear visual guide to the whole system, perfect for smooth reviews or handoffs.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 39 to 39 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is code_diver?

Code Diver scans your codebase to build an interactive graph blending human-written semantic annotations from .dive markdown files with static signals like imports and calls in Rust, TypeScript/JavaScript, Python, and Go. Run it via CLI on any project dir—`cargo run -- `—and get a live-updating web UI at localhost:4000 plus an MCP server for agent queries, perfect for reviewing sprawling LLM-generated code where questions like "can LLMs write code?" or "how LLMs write code" meet reality. It watches changes, overlays git diffs, and highlights coverage gaps, turning code divergence into navigable maps.

Why is it gaining traction?

Unlike static analyzers or basic call graphs, it fuses intent from .dive overviews/modules with heuristic edges, pruning for clarity while supporting real-time reloads and VS Code jumps. Devs dig the explorable viz—zoom into components, trace relations, isolate cycles or folders—plus MCP tools for "code_graph_overview" or "relationship_trace," making LLM code review (parallel code, GPU kernels, even books?) feel structured. Low overhead, no cloud, just local Rust speed.

Who should use this?

AI-assisted devs auditing LLM outputs for handoffs, backend teams mapping monorepos with code diversity, or agent builders querying graphs via MCP. Ideal for Rust/TS/Python/Go projects where semantic tags clarify "do LLMs write like humans?" in massive diffs.

Verdict

Grab it for LLM code review workflows—quick setup, sharp UI—but 39 stars and 1.0% credibility signal early days; docs are solid, but expect rough edges in edge cases. Worth a spin if code diver empire intrigues you.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.