Calliope CLI

The Muse of Digital Eloquence

Multi-model AI agent CLI with autonomous loops, project memory, and advanced tooling. Use Claude, Gemini, GPT, and more from a single elegant interface.

 ██████╗ █████╗ ██╗     ██╗     ██╗ ██████╗ ██████╗ ███████╗
██╔════╝██╔══██╗██║     ██║     ██║██╔═══██╗██╔══██╗██╔════╝
██║     ███████║██║     ██║     ██║██║   ██║██████╔╝█████╗
██║     ██╔══██║██║     ██║     ██║██║   ██║██╔═══╝ ██╔══╝
╚██████╗██║  ██║███████╗███████╗██║╚██████╔╝██║     ███████╗
 ╚═════╝╚═╝  ╚═╝╚══════╝╚══════╝╚═╝ ╚═════╝ ╚═╝     ╚══════╝

        The Muse of Digital Eloquence

Installation

npm install -g @calliopelabs/cli

Quick Start

# Run Calliope (first run will prompt for setup)
calliope

# Or set API key via environment
export ANTHROPIC_API_KEY=sk-ant-...
calliope

# Run in god mode (no permission prompts)
calliope -g
calliope --god-mode

Features

Multi-Model Support

Switch between 12+ providers on the fly:

/provider anthropic    # Use Claude
/provider google       # Use Gemini
/provider openai       # Use GPT
/provider mistral      # Use Mistral
/provider ollama       # Use local models
/provider openrouter   # Use any model via OpenRouter

Supported Providers:

• Anthropic (Claude): Sonnet 4, Opus 4, Haiku
• OpenAI: GPT-4o, o1, o3-mini
• Google: Gemini 2.0 Flash, 1.5 Pro
• Mistral: Large, Small
• Groq: Llama 3.3, Mixtral
• Cerebras: Llama 3.3
• Fireworks: Llama models
• xAI: Grok
• Ollama: Local models
• OpenRouter: 100+ models
• GitHub Models: Via Azure
• DeepSeek: R1, V3

🔒 Scope Management

Control which directories your AI can access for enhanced security:

/scope add ./src              # Grant access to src directory
/scope add ./tests            # Add tests directory
/scope list                   # View current scope
/scope remove ./src           # Revoke access
/scope reset                  # Clear all scope restrictions

Benefits:

• 🔒 Security - Prevent accidental modifications outside project areas
• 🎯 Focus - Keep AI operations within relevant directories
• ⚡ Performance - Reduce unnecessary file scanning
• 📋 Audit trail - Track which directories are in scope

🔁 Session Management & Continuity

Seamless continuity between sessions:

# On startup, Calliope detects previous sessions:
╭─ 🎭 Calliope v0.6.11
│  Found previous session (2 hours ago)
│    • 12 messages, 2 TODOs pending
╰─ [R]esume  [N]ew session

# Commands
/session list          # List all sessions
/session info          # Current session details
/history               # View conversation history
/export [file.md]      # Export conversation to markdown

🔖 Bookmarks & Navigation

Mark and return to important moments in your conversation:

/bookmark "Got auth working"    # Create a bookmark
/bookmarks                      # List all bookmarks
/goto bookmark-1                # Jump to that point in history
/bookmark delete 1              # Remove a bookmark

Use cases:

• Mark successful implementations
• Save decision points
• Quick navigation in long conversations

📋 Templates

Save and reuse common prompts and workflows:

/template save code-review "Review this code for bugs, performance, and best practices"
/template list                  # View all templates
/template use code-review       # Use a saved template
/template delete code-review    # Remove a template

Built-in templates include:

• Code review
• Bug investigation
• Refactoring plans
• Testing strategies

↩️ Undo/Redo

Navigate conversation history with full state management:

/undo                  # Undo last exchange (message + all responses)
/redo                  # Redo previously undone exchange
/history               # View full conversation timeline

Features:

• Preserves entire conversation state
• Works across tool executions
• Unlimited undo/redo stack
• Shows available undo/redo count

💰 Cost Tracking

Monitor API usage and costs across sessions:

# Automatic tracking in status bar:
# anthropic:claude-sonnet-4 │ 5.2K/200K │ 12.5K used │ $0.45

/cost                  # Detailed cost breakdown

Tracks:

• Current session cost
• Total cost across all sessions
• Cost per provider
• Token usage (input/output)
• Persists across sessions

⌨️ Advanced Input Experience

Enhanced input with modern conveniences:

• Tab completion - Slash commands and file paths autocomplete
• History navigation - Up/down arrows to browse previous inputs
• Multi-line editing - Shift+Enter for new lines (if supported)
• Path suggestions - Real-time file path completion
• Command hints - Footer shows available options

Example:

> /boo[TAB]           → /bookmark
> /scope add ./[TAB]  → ./src/  ./tests/  ./docs/
> [Up Arrow]          → Previous command

🚀 Parallel Tool Execution

Automatically executes independent tool calls in parallel for 2-5x speedup:

# When AI needs to read 3 files:
Sequential: Read file1 → wait → Read file2 → wait → Read file3
Parallel:   Read file1, file2, file3 → all at once!

# Dependency-aware:
- Detects when tools depend on each other
- Executes in optimal order
- Shows parallel execution progress

⚡ Autonomous Loops

Enable continuous execution for complex multi-step tasks:

/loop on               # Enable autonomous loops
/loop off              # Disable
/loop 10               # Set max iterations

# The AI will:
# 1. Plan the approach
# 2. Execute tools
# 3. Evaluate results
# 4. Continue until goal achieved or max iterations

🧠 Project Memory (CALLIOPE.md)

Persistent memory across sessions using markdown files:

/memory init           # Create CALLIOPE.md
/memory add context "This is a TypeScript project"
/memory add preference "Use functional components"
/memory show           # View project memory
/memory global         # View global preferences

Calliope automatically loads context from standard files:

• CALLIOPE.md - Project memory
• CLAUDE.md - Claude context
• README.md, SPEC.md, TODO.md
• ARCHITECTURE.md, DESIGN.md, NOTES.md
• .cursorrules, .github/copilot-instructions.md

🛠️ Tools

Built-in tools for autonomous operation:

🐳 Sandboxed Code Execution

Execute code safely in Docker containers:

The execute_code tool automatically:
- Uses Docker when available (recommended)
- Falls back to local execution if needed
- Shows [sandboxed] or [unsandboxed] status
- Enforces resource limits and timeouts

🌐 MCP Server Support

Connect external tools via Model Context Protocol:

/mcp add https://mcp-server.example.com
/mcp list              # Show connected servers
/mcp tools             # List available tools from all servers
/mcp refresh           # Reconnect all servers

📦 Agent Skills (AgentSkills.io)

Install reusable skills from the community registry:

/skills add git-workflow
/skills add code-review
/skills list           # Show installed skills
/skills info <name>    # Skill details

🌲 Conversation Branching

Fork conversations to explore different approaches:

/branch new experiment "Try approach B"
/branch list           # Show all branches
/branch switch experiment
/branch delete experiment

🎨 Themes & Personas

Customize the visual style and AI personality:

# Themes
/theme list            # Show available themes
/theme monokai         # Set theme
/theme                 # Cycle through themes

# Personas
/persona professional  # Professional, concise responses
/persona minimal       # Minimal output, just essentials
/persona calliope      # Poetic, elegant responses (default)

Available themes: default, light, monokai, nord, minimal

🪝 Hooks System

Run custom scripts before/after tool execution:

/hooks init            # Initialize default hooks
/hooks list            # Show configured hooks
/hooks add pre-shell "echo Running: $CALLIOPE_COMMAND"

Hook events: pre-tool, post-tool, pre-shell, post-shell, pre-write, post-write, session-start, session-end

📊 Context Management & Warnings

Smart context tracking with proactive warnings:

⚠️  Context at 85% capacity (170K/200K tokens)
   Consider: /summarize compact | /clear | shorter messages

🚨 Context at 95% capacity! (190K/200K tokens)
   Action required: /summarize compact | /clear

Commands:

/summarize context     # View conversation summary
/summarize compact     # Compress context to fit limits
/clear                 # Clear conversation

🔍 Smart Search

Find anything in your conversation or project:

/search error handling    # Fuzzy search through history
/find auth               # Find files in project

⚙️ Modes

Three distinct operating modes for different workflows:

/mode plan             # Analyze and plan, don't execute
/mode hybrid           # Plan then execute (default)
/mode work             # Execute directly without planning

🛡️ Risk Assessment & Safety

Built-in safety for potentially dangerous operations:

• Automatic risk classification (low/medium/high/critical)
• Permission prompts for risky operations
• God mode (-g) to bypass prompts for trusted tasks
• Detailed explanations of risks
• Scope restrictions for file access

🚄 Streaming & Real-time Feedback

Live response rendering for immediate feedback:

• Token-by-token streaming for text responses
• Live tool execution status with progress indicators
• Visual diffs for file changes
• Parallel tool execution visualization
• Thinking process display

Commands Reference

Core Commands

/help, /h              # Show all commands
/exit, /quit, /q       # Exit Calliope
/clear, /c             # Clear conversation
/status, /s            # Show current status
/config                # Show configuration
/debug [on|off]        # Toggle debug mode

Model & Provider

/provider, /p [name]   # Switch AI provider
/model, /m [name]      # Set model (interactive if no name)
/models                # Browse available models
/persona [name]        # Switch persona (calliope/professional/minimal)

Modes & Settings

/mode [plan|hybrid|work]  # Switch operating mode
/loop [on|off|N]       # Toggle/configure autonomous mode
/confirm [on|off]      # Toggle risky operation confirmation
/theme [name]          # Switch theme

Scope Management

/scope add <path>      # Add directory to scope
/scope remove <path>   # Remove directory from scope
/scope list            # Show current scope
/scope reset           # Clear all scope restrictions

Session & History

/session list          # List all sessions
/session info          # Current session details
/history [search]      # View/search conversation history
/bookmark [name]       # Create bookmark
/bookmarks             # List bookmarks
/goto <id>             # Jump to bookmark
/undo                  # Undo last exchange
/redo                  # Redo exchange

Templates & Memory

/template save <name> <prompt>  # Save template
/template list         # View all templates
/template use <name>   # Use template
/template delete <name> # Remove template

/memory init           # Initialize project memory
/memory add <text>     # Add to memory
/memory show           # View project memory
/memory global         # View global memory

Context & Search

/summarize [context|compact]  # Summarize/compress conversation
/search <query>        # Search conversation history
/find <pattern>        # Fuzzy file search in project
/branch [new|switch|list]  # Conversation branches

Tools & Extensions

/mcp [list|add|remove|tools]  # MCP server management
/skills [list|add|remove|info] # Agent skills
/hooks [list|add|init] # Pre/post tool hooks

Tasks & Planning

/todo [add|done|work|clear|list]  # Manage TODOs
/plans [list|view|rerun]          # View and rerun plans

Utility

/cost                  # Show cost breakdown
/copy                  # Copy last response
/export [file.md]      # Export conversation to file
/upgrade               # Check for updates
/set <key> <value>     # Set runtime config (e.g., maxIterations)
/context               # Show loaded project context

Environment Variables

# API Keys
ANTHROPIC_API_KEY=sk-ant-...
OPENAI_API_KEY=sk-...
GOOGLE_API_KEY=...
MISTRAL_API_KEY=...
GROQ_API_KEY=...
XAI_API_KEY=...
CEREBRAS_API_KEY=...
FIREWORKS_API_KEY=...
DEEPSEEK_API_KEY=...
OPENROUTER_API_KEY=...

# Configuration
CALLIOPE_PROVIDER=anthropic
CALLIOPE_MODEL=claude-sonnet-4-20250514
CALLIOPE_GOD_MODE=false
CALLIOPE_DEBUG=false

# Ollama
OLLAMA_BASE_URL=http://localhost:11434

# LiteLLM
LITELLM_BASE_URL=http://localhost:4000
LITELLM_API_KEY=...

Configuration

Config is stored in ~/.config/calliope/config.json (or platform equivalent).

# Show config location
calliope --config

# Reset config
calliope --reset

# Force setup wizard
calliope --setup

Sample config.json

{
  "provider": "anthropic",
  "model": "claude-sonnet-4-20250514",
  "apiKeys": {
    "anthropic": "sk-ant-...",
    "openai": "sk-...",
    "google": "..."
  },
  "theme": "default",
  "persona": "calliope",
  "godMode": false,
  "maxIterations": 25,
  "scope": ["./src", "./tests"]
}

Profiles

Save and switch between configurations:

/profile save work     # Save current settings as "work"
/profile work          # Load "work" profile
/profile delete work   # Delete profile
/profile list          # Show all profiles

Built-in profiles: fast, smart, cheap, local

Keyboard Shortcuts

File References

Reference files directly in your messages:

@filename.ts           # Include file content
./path/to/file         # Relative path
/absolute/path/file    # Absolute path

Images are automatically detected and sent for vision-capable models.

Security & Best Practices

API Key Storage

• Keys stored in ~/.config/calliope/config.json
• Prefer environment variables for better security
• Never commit config to version control

Tool Execution Safety

• Path traversal protection
• Sandboxed code execution via Docker
• Timeout limits on all commands
• Risk assessment for dangerous operations
• Scope restrictions for file access
• Hook system for custom validation

Best Practices

1. Use /scope to limit directory access
2. Review tool calls before execution (unless in god mode)
3. Set iteration limits on autonomous loops
4. Work from project directories, not system directories
5. Regular /summarize to manage context
6. Use bookmarks for important moments

Development

# Clone and setup
git clone https://github.com/calliopeai/calliope-cli.git
cd calliope-cli
npm install

# Build
npm run build

# Run locally
node dist/bin.js

# Run tests
npm test

# Watch mode for development
npm run dev

Architecture

• TypeScript - Full type safety throughout
• Ink (React) - Declarative CLI rendering
• Vitest - Fast unit testing framework
• Multi-provider - Unified interface for 12+ LLM providers
• Modular design - Clean separation of concerns
• Parallel execution - Dependency-aware tool execution
• Error boundaries - Graceful crash recovery

Troubleshooting

"No API keys configured"

• Run calliope --setup to configure keys
• Or set environment variables (see above)

"Empty response from API"

• Check API key validity
• Verify account credits/quota
• Try a different provider

Context limit warnings

• Use /summarize compact to compress
• Or /clear to start fresh
• Consider shorter messages

"Access denied" for file operations

• Operations restricted to cwd and home directory
• Check /scope list for current restrictions
• Use /scope add <path> to grant access

Docker not available for sandbox

• Install Docker for safer code execution
• Without Docker, code runs locally (less safe)

Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

License

MIT License - see LICENSE

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Twitter: @calliopelabs

Acknowledgments

Built with love by the Calliope team. Special thanks to:

• Anthropic for Claude
• OpenAI for GPT
• Google for Gemini
• The open-source community
• All contributors and users

May your code be elegant, your execution swift, and your automation poetic. 🎭✨