ai-terminal
AI-Terminal is an advanced AI-powered CLI that enhances terminal workflows through AI-driven automation and optimization.
Links
README
From the repo.
AI-Terminal
🇺🇸 English | 🇨🇳 中文 | 🇯🇵 日本語
Your intelligent command-line companion that transforms terminal workflows with AI-powered automation
AI-Terminal is an advanced command-line interface enhanced with artificial intelligence to help developers work smarter, not harder. It provides context-aware assistance, intelligent automation, and powerful code generation capabilities right in your terminal.
✨ Key Features
🤖 Interactive Coding Sessions
- Multi-mode Interface: Switch between chat (
/ask), coding (/coding), and execution (/exec) modes - File Context Management: Load files with
/addcommand for context-aware operations - Real-time Code Editing: Apply AI-generated code changes directly to your files
- Auto-commit Integration: Automatically commit changes after code generation
📝 Intelligent Commit Automation
- Multi-stage Analysis: Code review → Title generation → Prefix determination → Message creation
- Language Support: Generate commit messages in English, Chinese, Japanese, Portuguese, and Brazilian Portuguese
- Conventional Commits: Automatic prefix detection (feat, fix, docs, style, refactor, test, chore)
- Flexible Attribution: Configurable author/committer attribution for AI-generated changes
🔧 Smart Command Execution
- OS-Aware Suggestions: Commands tailored to your operating system and architecture
- Safety First: Built-in confirmation prompts and
[noexec]responses for unsafe operations - Context Integration: Leverages current user, OS info, and directory context
💬 Advanced File & Context Management
- Pattern Matching: Support for glob patterns and directory recursion
- Remote Content: Load and analyze content from URLs
- Session Persistence: Save and restore conversation contexts across sessions
- Batch Operations: Process multiple files and directories efficiently
🚀 Quick Start
Prerequisites
- Go 1.22.0 or later (if building from source)
- Git installed and configured
- Terminal/Command Prompt access
- API access to supported AI providers
Installation
Homebrew (Recommended):
brew install coding-hui/tap/ai-terminal
Download Pre-built Binaries: Visit our [releases page][releases] to download binaries for:
- Linux (x86_64, arm64)
- macOS (Intel, Apple Silicon)
- Windows
Build from Source:
git clone https://github.com/coding-hui/ai-terminal.git
cd ai-terminal
make build
Initial Configuration
ai configure
This interactive setup will guide you through API key configuration and personal preferences.
📖 Usage Examples
Interactive Coding Session
# Start interactive coding
ai coder
# In the session:
/add *.go # Add all Go files
/add https://example.com/api-spec.json # Add remote content
/coding "Refactor these functions to use interfaces"
/commit # Auto-commit the changes
Smart Commit Messages
# Stage your changes
git add .
# Generate commit with specific language and prefix
ai commit --lang zh-cn --prefix fix
# Preview before committing
ai commit --preview --no-confirm
# Amend previous commit
ai commit --amend
Command Assistance & Execution
# Get command suggestions
ai exec "find all recently modified Go files and count lines"
# Execute with auto-confirmation
ai exec --yes "list all Docker containers and their status"
# Safe rejection for destructive operations
ai exec "rm -rf /" # Returns [noexec] with reason
File Analysis & Q&A
# Load files and ask questions
ai coder
/add package.json
/add src/
/ask "What dependencies are outdated and what are the alternatives?"
Design & Architecture
ai coder
/add current-architecture.md
/design "Design a microservices architecture for this monolith"
⚙️ Configuration
API Setup
Configure multiple AI providers in your config file:
apis:
openai:
api-key: sk-...
base-url: https://api.openai.com/v1
localai:
api-key: ""
base-url: http://localhost:8080/v1
Auto-coder Settings
auto-coder:
coding-model: gpt-4
design-model: gpt-4
auto-commit: true
commit-prefix: feat
coding-fences: ["```", "```"]
🏗️ Advanced Features
Session Management
# Save current context
/save session-name
# List available sessions
/sessions
# Clear current conversation
/clear
Diff and Apply Operations
# View formatted diffs
/diff
# Apply edit blocks directly
/apply "
<<<<<<< SEARCH
old code
Collected info
- ★ 28 stars
- ⎇ 6 forks
- Language: Go
- Source updated: 8/2/2026
Config for your environment
Replace {MCP_ENDPOINT_URL} with this MCP’s endpoint URL (from its repo or docs above). No API key — you connect directly.
Tool
OS
Config file: ~/.cursor/mcp.json
{
"mcpServers": {
"mcp-server": {
"url": "{MCP_ENDPOINT_URL}"
}
}
}Paste into mcpServers in the config file. Restart Cursor after saving.
If this MCP is also published on mcpchannel.ai, you can subscribe from Browse and use the gateway config there instead.