vtcode  by vinhnx

Semantic coding agent for terminal development

Created 3 months ago
275 stars

Top 94.1% on SourcePulse

GitHubView on GitHub
Project Summary

VT Code is a Rust-based terminal coding agent designed to bring semantic code intelligence and AI assistance directly into the command line. It targets developers seeking to enhance their terminal-based workflows by integrating advanced language understanding and LLM capabilities. The primary benefit is a powerful, AI-driven coding assistant that operates within the familiar terminal environment, supporting multiple AI providers and offering robust security features.

How It Works

VT Code leverages Rust for performance and utilizes Tree-sitter parsers to achieve deep semantic understanding of code across multiple languages, including Rust, Python, JavaScript/TypeScript, Go, Java, and Swift. It integrates seamlessly with a wide array of LLM providers (e.g., OpenAI, Anthropic, Gemini, Ollama) featuring automatic failover for reliability. An Agent Client Protocol (ACP) enables integration with code editors like Zed, while advanced context management ensures efficient token usage.

Quick Start & Requirements

Installation is straightforward via native scripts or package managers:

  • macOS/Linux: curl -fsSL https://raw.githubusercontent.com/vinhnx/vtcode/main/scripts/install.sh | bash
  • Windows (PowerShell): irm https://raw.githubusercontent.com/vinhnx/vtcode/main/scripts/install.ps1 | iex
  • Package Managers: brew install vtcode, cargo install vtcode, npm install -g @vinhnx/vtcode

A prerequisite is setting up your chosen LLM provider's API key via environment variables (e.g., export OPENAI_API_KEY="sk-..."). Further configuration details and troubleshooting guides are available in the linked documentation.

Highlighted Details

  • Security First: Implements a defense-in-depth model with execution policies, workspace isolation, sandbox integration, human-in-the-loop approval, and audit trails.
  • Multi-Provider AI: Supports numerous LLMs including OpenAI, Anthropic, Google Gemini, xAI, DeepSeek, OpenRouter, Ollama (local), and LM Studio (local).
  • Code Intelligence: Utilizes Tree-sitter for accurate semantic code analysis.
  • Editor Integration: Native support for Zed IDE via ACP and compatibility with other VS Code-like editors. Also available as a VS Code extension.
  • Lifecycle Hooks: Enables custom shell commands to be executed in response to agent events for automation and context enrichment.
  • TUI Interface: Provides a rich terminal user interface with real-time streaming output.

Maintenance & Community

VT Code is described as a "passion project" developed in the creator's free time. Support is encouraged via BuyMeACoffee. Specific community channels like Discord or Slack are not detailed in the provided README snippet.

Licensing & Compatibility

The project is licensed under the permissive MIT License, which generally allows for commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

As a passion project, development pace and long-term maintenance may differ from commercially backed initiatives. While extensive, specific setup instructions for all supported LLM providers and ACP integrations require consulting external documentation. The existence of a separate VS Code extension might indicate distinct feature sets or development tracks.

Health Check
Last Commit

1 day ago

Responsiveness

Inactive

Pull Requests (30d)
23
Issues (30d)
59
Star History
69 stars in the last 30 days

Explore Similar Projects

Starred by Alex Yu Alex Yu(Research Scientist at OpenAI; Cofounder of Luma AI), Will Brown Will Brown(Research Lead at Prime Intellect), and
7 more.

avante.nvim by yetone

0.3%
17k
Neovim plugin emulating Cursor AI IDE for AI-driven code assistance
Created 1 year ago
Updated 2 days ago
Starred by Gabriel Almeida Gabriel Almeida(Cofounder of Langflow) and Chip Huyen Chip Huyen(Author of "AI Engineering", "Designing Machine Learning Systems").

serena by oraios

1.5%
17k
Coding agent toolkit for LLMs, enabling codebase interaction
Created 8 months ago
Updated 1 day ago
Feedback? Help us improve.