johunsang

johunsang / semble_rs

Public

Fast, AI-agent-native code search in Rust — hybrid BM25 + semantic, Tree-sitter AST chunking, dependency & impact analysis. Drop-in replacement for grep/cat/read/ls in Claude Code, Codex, Cursor, Aider, OpenHands.

github.comjohunsangsemble_rs ai-agent bm25 claude-code cli code-search
33
5
100% credibility
Found May 14, 2026 at 71 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

semble_rs is a fast command-line tool for hybrid keyword and semantic code search with dependency analysis, optimized for AI coding agents to replace traditional file reading workflows.

How It Works

1
📖 Discover the smart code finder

You learn about a helpful tool that lets AI assistants quickly find exactly the right parts of code without reading everything.

2
🛠️ Set it up simply

You follow easy steps to get the tool ready on your computer—it downloads what it needs automatically.

3
📁 Pick your project

You point the tool at your folder full of code files.

4
🔍 Ask for code

You describe what you want in plain words, like 'user login function', and it searches smartly.

5
📋 Get perfect snippets

You see ranked bits of code with exact line numbers—no need to open full files.

6
🔗 Explore connections

You check which files depend on each other or what changes might affect.

🚀 AI works faster

Your AI helper now explores code super quick, saving time and effort every session.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 71 to 33 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 semble_rs?

Rust CLI for ai-agent-native code search across Rust, Python, JS/TS, Go, Java, C/C++, and more. Hybrid BM25 + semantic search with AST chunking delivers ranked snippets, line numbers, and match highlights in one call – replacing grep/cat/read/ls chains in Claude Code, Codex, Cursor, Aider, OpenHands. Agents get precise code context without reading full files, slashing tokens up to 93%.

Why is it gaining traction?

Pure Rust binary runs fast anywhere, no pip/venv hassle like the Python original. Adds dependency analysis (`deps file .`), impact analysis (`impact file .`), and Unicode BM25 for Korean/CJK code – plus smart outputs like `--compact` or `--outline` for minimal tokens. Outperforms grep in semantic queries while matching size, ideal for fast github search in agent loops.

Who should use this?

AI agent devs tuning Claude Code/Cursor sessions; Aider/OpenHands users needing code retrieval; refactoring teams querying deps/impact before changes. Frontend/backend folks exploring unfamiliar repos via natural language or symbols.

Verdict

Strong agent tool with excellent docs and token savings – try `cargo install` for fast github actions/runners. 33 stars and 1.0% credibility flag early maturity; lacks broad testing, but MIT license invites contributions.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.