Skip to content

Codeilus

Status: Pre-release (v0.1.0-alpha)

All core features functional. Some narrative types may fall back to placeholders if LLM times out.

Turn any codebase into an interactive learning experience.

  • Get Started in 30 Seconds


    git clone https://github.com/mbaneshi/codeilus.git
    cd codeilus && cargo build --release
    ./target/release/codeilus analyze ./any-repo
    ./target/release/codeilus serve
    # Open http://localhost:4174
    

    Installation

  • Guided Learning Path


    Auto-generated chapters ordered by dependency. Quizzes, XP, badges, and progress tracking.

    Learning Path

  • Interactive Graph Explorer


    Navigate the knowledge graph with colored communities, call chains, and impact analysis.

    Graph Explorer

  • AI-Powered Understanding


    8 types of pre-generated narratives plus streaming Q&A powered by Claude Code.

    Ask AI


What Codeilus Does

Before Codeilus After Codeilus
Grep around, hope you find the right file Guided reading order: "read these 3 files, understand 80%"
Stare at a class and guess what it does LLM-generated explanations with graph context
No idea how modules connect Interactive architecture diagram from real data
Onboarding takes weeks Gamified learning path: chapters, quizzes, progress
README is the only overview 8 types of pre-generated narrative content

Two Modes, One Engine

Interactive Local Server

./target/release/codeilus analyze ./any-repo
# Analyzing... 342 files, 2847 symbols, 12 communities
# Narrating... 8 sections generated

./target/release/codeilus serve
# Open http://localhost:4174

Full SvelteKit 5 UI with guided learning, graph explorer, metrics dashboard, architecture diagrams, streaming Q&A, and gamification.

Static Page Publishing

codeilus harvest --trending
codeilus export --all-harvested --output ./output
codeilus deploy ./output --cloudflare

Self-contained HTML pages (~300KB) that load in under 1 second. Daily automation via GitHub Actions.


Architecture at a Glance

codeilus (single binary, ~15MB)
+-- Rust backend (Axum HTTP + WebSocket)
+-- SQLite database (WAL mode, zero config)
+-- SvelteKit 5 frontend (embedded via rust-embed)
+-- Claude Code integration (subprocess, stream-json)

No Docker. No PostgreSQL. No Redis. No Node.js runtime. Just one binary + Claude Code.

Full architecture