mandark  by hrishioa

AI coding assistant for code analysis and manipulation

Created 1 year ago
251 stars

Top 99.8% on SourcePulse

GitHubView on GitHub
1 Expert Loves This Project
Project Summary

Summary

Mandark is a lightweight, command-line AI intern designed to automate coding tasks and assist developers. It targets individual developers and power users seeking to streamline workflows by integrating with various AI models for code analysis, generation, and modification directly from their terminal. The primary benefit is enhanced productivity through AI-powered code interaction without leaving the command line.

How It Works

This tool leverages Node.js's npx for execution, allowing it to run without a formal installation. Mandark processes codebases by accepting file paths or GitHub repository URLs. It supports multiple large language models, including Claude 3.5 Sonnet/Haiku, GPT-4o-mini/turbo, and Llama405b, enabling flexible AI backend choices. Core functionalities include asking questions about code, copying code snippets, piping code to other processes, and even editing/creating files, all while providing token and cost estimations before execution.

Quick Start & Requirements

  • Primary command: npx mandark <folder or file> or npx mandark <mode> ...
  • Prerequisites: Node.js and npm (for npx). Access to supported AI models (API keys may be required depending on the model provider, though not explicitly stated).
  • Setup: Minimal, as it runs via npx.
  • Relevant links: Project repository: https://github.com/hrishioa/mandark

Highlighted Details

  • Multi-modal AI Integration: Supports a range of models including Claude 3.5, GPT-4 variants, and Llama405b.
  • Versatile Command-Line Modes: Offers ask for code Q&A, copy for clipboard integration, and pipe for stdout redirection.
  • Code Management: Capable of editing/creating multiple files, verifying diffs, and optionally including import statements (-a flag).
  • Efficiency Features: Includes token and cost estimation prior to execution, and a compact ~80kb footprint.

Maintenance & Community

Contributions are welcomed via Pull Requests, with TODO.md suggested as a starting point. No specific community channels (like Discord/Slack) or notable contributors/sponsorships are detailed in the provided README.

Licensing & Compatibility

The license type is not specified in the provided README. Compatibility for commercial use or linking with closed-source projects is therefore undetermined.

Limitations & Caveats

The project is described as "bog-simple" and an "AI intern," suggesting it may not handle highly complex, enterprise-level refactoring or architectural design tasks without significant oversight. The lack of explicit licensing information presents an adoption blocker for users requiring clear legal terms.

Health Check
Last Commit

6 months ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
4 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.