darshil3011

darshil3011 / codedrift

Public

Your coding agent spends 90% of its tokens finding code, not writing it. Reduce your token usage by 50x with CodeDrift !

34
7
100% credibility
Found Apr 21, 2026 at 34 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

CodeDrift indexes codebases locally to provide AI coding agents with efficient search, resolution, and diff tools for functions, classes, calls, and more, drastically reducing token consumption.

How It Works

1
🔍 Discover CodeDrift

You learn about a handy helper that lets your AI coding buddy find exactly the right bits of code super quickly, without wasting time or money rereading everything.

2
📥 Add it to your project

You bring CodeDrift into your project's main folder with a simple download and setup.

3
⚡ Scan your codebase

You press go once, and it quickly reads through all your files to build a smart map of functions, classes, and connections – now everything is searchable in a snap.

4
🔗 Link to your AI assistant

You connect it to your AI coding tool so it knows to use these smart lookup tricks instead of hunting blindly.

5
🔄 Keep it fresh automatically

It updates itself whenever you change your code, staying ready without extra work from you.

6
💰 Watch your AI work smarter

During coding sessions, your AI grabs precise code snippets, diffs for changes, and even recalls past helpful files – feeling lightning-fast and efficient.

🎉 Save big on AI costs

You measure huge savings in AI usage, code faster with less waste, and enjoy a sidekick that truly understands your project.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 34 to 34 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 codedrift?

CodeDrift indexes your Python, JavaScript, Go, or Rust codebase using tree-sitter parsers into a local SQLite database with full-text search, letting coding agents query symbols, callers, and diffs without wasteful greps or full file reads. It slashes token usage by up to 50x for tools like Claude Code by delivering precise context—source, tests, git history—and tracking session changes for delta-only re-reads. Install via pip, run `codedrift init` to index, and expose MCP tools for seamless agent integration.

Why is it gaining traction?

In coding agents comparison and ranking tests, it shines on coding agent benchmark metrics by eliminating the 90% token waste on file hunting, with a CLI benchmark script proving savings on your sessions. Features like cross-session memory recall via embeddings and git commit hooks for auto-updates make agents smarter over time, standing out from basic github coding ai tools. MCP support hooks directly into Claude Code, bypassing native grep/glob chains for faster, cheaper responses.

Who should use this?

Devs running Claude Code or local LLM coding agents on mid-size repos, especially backend teams building Python/Go services tired of token burn on github coding agent jobs. Open source maintainers evaluating coding agent open source options or coding agent pi setups for offline use, and anyone in coding agent github workflows needing symbol resolution without intellij plugins.

Verdict

Early alpha with 34 stars and 1.0% credibility score—docs are solid, setup takes minutes, but expect bugs in edge cases; test on your repo first. Worth trying if you're optimizing coding agent pricing and skills in agent-heavy flows.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.