terminal-copilot-cli 0.1.3

A CLI tool that wraps shell commands with diagnostics and AI-ready workflows.

Terminal Copilot

A CLI tool that wraps shell commands with diagnostics and an AI-ready workflow foundation.

Installation

uv sync

Setup

Set your Gemini API key using one of these methods:

Option 1: Environment Variable

export GEMINI_API_KEY=your-api-key-here

Option 2: .env File

Create a .env file in the project root:

cp .env.example .env

Then edit .env and add your API key:

GEMINI_API_KEY=your-actual-api-key-here

Get your API key from Google AI Studio.

Usage

terminal-copilot run "<command>"

Examples

# Run a simple command
terminal-copilot run "echo hello world"

# Run a command that fails
terminal-copilot run "ls /nonexistent"

# Run with plugin detection
terminal-copilot run "git status"
terminal-copilot run "npm install"
terminal-copilot run "docker ps"

Output

The tool displays:

• The command that was executed
• The matching plugin (npm, docker, git, or unknown)
• Exit code (green for success, red for failure)
• Execution time
• stdout and stderr in separate panels

Configuration

Create ~/.terminal-copilot/config.yaml:

provider: gemini
auto_execute: false
plugins:
  - npm
  - docker
  - git

Project Structure

terminal_copilot/
    __init__.py    # Package init
    cli.py         # Typer CLI entry point
    runner.py      # Shell command execution
    config.py      # YAML config loader
    plugins.py     # Plugin interface (npm, docker, git)
    workflow.py    # LangGraph workflow skeleton
    models.py      # Pydantic models

pyproject.toml
README.md

Tech Stack

• Python 3.12
• Typer – CLI framework
• Rich – Terminal UI
• Pydantic v2 – Data validation
• PyYAML – Config parsing
• LangGraph – Workflow graph

Milestone 5: Interactive Repair Loop

• Failure diagnosis with suggested commands
• Interactive fix selection prompt
• Execute chosen fix command
• Re-investigation on fix failure
• Loop until fix succeeds or user cancels
• Available in both run and explain commands

Interactive Repair Flow

When a command fails, Terminal Copilot now guides you through an interactive repair process:

$ terminal-copilot run "npm install"

# ... command output ...
# Command fails

🔍 Root Cause

Node version mismatch in package.json engines requirement.

💡 Suggested Commands

1. corepack enable
2. pnpm install  
3. Cancel

> 1
Executing: corepack enable
# ... fix attempt output ...
✓ Fix succeeded!

How It Works

1. Failure: Command executes and fails
2. Diagnosis: AI analyzes the failure and provides root cause + suggested fixes
3. Choose Fix: You select a suggested command to try
4. Execute: The chosen fix is executed
5. Succeeded?:
   • Yes → Done! Returns success
   • No → Re-investigates using the new output, looping back to step 3

This creates an interactive repair loop that continues until either:

• A fix succeeds
• You choose to cancel

Using with explain

The interactive repair loop is also available when analyzing a previously failed command:

terminal-copilot explain
# Shows diagnosis with suggested fixes
# You can select and execute a fix directly

Milestone 4: Project Health & Preflight Checks

• Preflight checks before command execution
• Plugin-specific health checks (npm, docker, git, python, rust, go)
• Warning system for potential issues
• Error blocking for preventing dangerous commands
• --skip-preflight flag to bypass checks
• Project type detection from marker files
• Environment validation against project requirements
• Missing dependency detection (node_modules, Cargo.lock)
• Intelligent warnings for proactive issue detection

Project Detection

Terminal Copilot can automatically detect your project type based on common marker files:

• package.json → Node
• Cargo.toml → Rust
• go.mod → Go
• requirements.txt or pyproject.toml → Python
• Dockerfile → Docker
• docker-compose.yml → Docker Compose

# Detect project type in current directory
terminal-copilot detect

# Detect project type in a specific directory
terminal-copilot detect --path /path/to/project

Environment Validation

Validate your environment against project requirements to catch compatibility issues:

• Node.js: Compares installed version against package.json engines.node requirement
• Python: Compares installed version against pyproject.toml python_requires requirement
• Docker: Checks if Docker daemon is running when Dockerfile/docker-compose.yml exists
• Dependencies: Detects missing dependency directories (node_modules, Cargo.lock)

# Validate environment compatibility
terminal-copilot validate

# Validate in a specific directory
terminal-copilot validate --path /path/to/project

Plugin Preflight Checks

Each plugin now provides preflight checks that run automatically before command execution. These checks detect common issues specific to each tool ecosystem:

NPM Plugin Example:

✓ package.json
✓ node v22.2.1
✓ npm 9.2.0
✓ lockfile
⚠ node_modules missing (run npm install)

Docker Plugin Example:

✓ Docker installed Docker version 24.0.0
✓ daemon running
⚠ Dockerfile
⚠ compose plugin missing

Available Plugin Checks

• npm: package.json, node version, npm version, lockfile, node_modules
• docker: Docker installed, daemon running, Dockerfile, compose plugin
• git: Git installed, in git repository
• rust: rustc, cargo, Cargo.toml
• go: Go installed, go.mod
• python: Python installed, virtual environment

Intelligent Warnings

Instead of waiting for failure, Terminal Copilot proactively detects potential issues before command execution:

• Git merge conflicts: Warns when your working directory has unresolved merge conflicts
• Docker daemon status: Warns when Docker is not running before executing docker commands
• Port conflicts: Warns when common development ports (3000, 3001, 4000, 5000, 5173, 8000, 8080, 9000) are already in use

Example warnings displayed before execution:

⚠ Git has merge conflicts in your working directory
  This build is likely to fail. Resolve conflicts before proceeding.

⚠ Docker daemon is not running
  Start Docker: 'systemctl start docker' or launch Docker Desktop

⚠ Port 3000 is already in use
  Another process is using port 3000. Check with 'lsof -i :3000' or kill the process.

Future Milestones

• Auto-fixing failed commands
• File scanning and project analysis
• Streaming output improvements
• Watchdog mode