toothfairy-cli 1.0.2

Command-line interface for ToothFairy AI API

ToothFairy AI CLI Tools

Command-line interfaces for interacting with ToothFairy AI agents. Available in both Python and Node.js versions with identical functionality.

Quick access: Use either tf or toothfairy commands!

📦 Installation

Python CLI (PyPI)

pip install toothfairy-cli

Node.js CLI (npm)

npm install -g @toothfairyai/cli

Both packages provide the same commands: toothfairy and tf

🚀 Getting Started

1. Configure your credentials

tf configure --api-key YOUR_KEY --workspace-id YOUR_WORKSPACE

2. Send a message to an agent

tf send "Hello, I need help" --agent-id YOUR_AGENT_ID

3. Search the knowledge hub

tf search "AI configuration help"

4. Explore your workspace

tf chats                    # List all conversations
tf config-show              # View current settings

✨ Features

• 💬 Agent Communication: Send messages to ToothFairy AI agents
• 🔍 Knowledge Hub Search: Search documents with advanced filtering
• 📋 Chat Management: List and view chat conversations
• ⚙️ Configuration: Flexible credential and settings management
• 🎨 Rich Output: Beautiful terminal interface with colors and tables
• 📊 Multiple Formats: JSON and text output for both interactive and scripted use
• 🔐 Secure: Safe credential storage and validation
• 📱 Cross-Platform: Works on Windows, macOS, and Linux

💬 Agent Communication Examples

Use Case	Command
Simple message	tf send "What are your hours?" --agent-id "info-agent"
With customer info	tf send "Schedule appointment" --agent-id "scheduler" --customer-info '{"name": "John"}'
Verbose output	tf send "Hello" --agent-id "agent-123" --verbose
JSON for scripting	tf send "Help" --agent-id "agent-123" --output json

Advanced Agent Communication

# Send message with detailed customer information
tf send "Check my insurance coverage" \
  --agent-id "insurance-agent" \
  --customer-info '{
    "name": "Sarah Johnson",
    "dateOfBirth": "1985-03-15",
    "insurance": {
      "provider": "BlueCross BlueShield",
      "policyNumber": "BC123456789"
    }
  }'

🔍 Knowledge Hub Search Examples

Use Case	Command
Basic search	tf search "AI agent configuration"
Filter by status	tf search "machine learning" --status published
Limit results	tf search "troubleshooting" --top-k 3
Topic filtering	tf search "automation" --topics "topic_123,topic_456"
Specific document	tf search "settings" --document-id "doc_550..."
Verbose details	tf search "deployment" --verbose
JSON output	tf search "API docs" --output json

Search Filtering Guide

Knowledge Hub search supports powerful filtering options:

• --status: Filter documents by 'published' or 'suspended' status
• --topics: Use topic IDs from ToothFairyAI (comma-separated)
• --document-id: Search within a specific document
• --top-k: Control number of results (1-50)
• --verbose: Show relevance scores and metadata

📋 Workspace Management Examples

Use Case	Command
List all chats	tf chats
View chat details	tf chat CHAT_ID
Show config	tf config-show
Detailed help	tf help-guide

⚙️ Configuration

Both CLIs support multiple configuration methods (in order of priority):

Configuration Methods

Method	Description	Example
Environment	Set environment variables	export TF_API_KEY=your_key
Config file	Use ~/.toothfairy/config.yml	api_key: your_key workspace_id: your_workspace
CLI arguments	Pass config file path	tf --config /path/to/config.yml send ...

Setting up configuration

Interactive Configuration

tf configure \
  --api-key "your-api-key" \
  --workspace-id "your-workspace-id"

# View current configuration
tf config-show

Environment Variables

# Required
export TF_API_KEY="your-api-key"
export TF_WORKSPACE_ID="your-workspace-id"

# Optional (defaults to production)
export TF_BASE_URL="https://api.toothfairyai.com"
export TF_AI_URL="https://ai.toothfairyai.com"

Config File Format

Create ~/.toothfairy/config.yml:

# Required
api_key: "your-api-key"
workspace_id: "your-workspace-id"

# Optional (defaults to production)
base_url: "https://api.toothfairyai.com"
ai_url: "https://ai.toothfairyai.com"

📊 Output Formats

Both CLIs support --output or -o flag:

• text (default): Pretty formatted tables and panels
• json: Raw JSON output for scripting

# Get JSON output for scripting
tf send "Hello" --agent-id "agent-123" --output json | jq '.agentResponse.contents.content'

# Search with JSON output
tf search "documentation" --output json | jq '.[].title'

🔧 Command Reference

Global Options

• --config, -c: Path to configuration file
• --verbose, -v: Enable verbose logging and API debugging

Commands

configure

Set up CLI credentials and configuration.

Options:

• --api-key: API key (required)
• --workspace-id: Workspace ID (required)
• --base-url: ToothFairy API base URL (optional)
• --ai-url: ToothFairy AI URL (optional)
• --config-path: Custom path to save config file

send <message>

Send a message to a ToothFairy AI agent.

Arguments:

• message: The message text to send

Options:

• --agent-id: Agent ID to send message to (required)
• --phone-number: Phone number for SMS channel (optional)
• --customer-id: Customer ID (optional)
• --provider-id: SMS provider ID (optional)
• --customer-info: Customer info as JSON string (optional)
• --output, -o: Output format (json|text)
• --verbose, -v: Show detailed response information

search <query>

Search for documents in the knowledge hub.

Arguments:

• query: The search query text

Options:

• --top-k, -k: Number of documents to retrieve (1-50, default: 10)
• --status: Filter by document status (published|suspended)
• --document-id: Search within specific document ID
• --topics: Comma-separated topic IDs to filter by
• --output, -o: Output format (json|text)
• --verbose, -v: Show detailed search information

chats

List all chats in the workspace.

Options:

• --output, -o: Output format (json|text)

chat <chat-id>

Get details of a specific chat.

Arguments:

• chat-id: The chat ID to retrieve

Options:

• --output, -o: Output format (json|text)

config-show

Display current configuration (with masked API key).

help-guide

Show detailed help with examples, common issues, and pro tips.

💡 Usage Examples

Basic Workflow

# Configure once
tf configure --api-key "tf-key-abc123" --workspace-id "workspace-456"

# Send messages to agents
tf send "I'd like to schedule a session" --agent-id "my-scheduler"

# Search for relevant information
tf search "appointment scheduling help"

# Manage conversations
tf chats
tf chat chat-abc123

Scripting Examples

#!/bin/bash

# Send message and extract agent response
RESPONSE=$(tf send "What are your hours?" \
  --agent-id "info-agent" \
  --output json)

# Extract the agent's text response
AGENT_TEXT=$(echo "$RESPONSE" | jq -r '.agentResponse.contents.content')
echo "Agent said: $AGENT_TEXT"

# Search documents and extract relevant content
SEARCH_RESULTS=$(tf search "office hours" --output json)
TOP_RESULT=$(echo "$SEARCH_RESULTS" | jq -r '.[0].raw_text')
echo "Found documentation: $TOP_RESULT"

# Get chat ID for follow-up
CHAT_ID=$(echo "$RESPONSE" | jq -r '.chatId')
echo "Chat ID: $CHAT_ID"

⚠️ Common Issues & Solutions

Issue	Solution
Configuration incomplete	Run: tf configure --api-key YOUR_KEY --workspace-id YOUR_WORKSPACE
No text response found	Use --verbose flag to see full response details
Agent not responding	Check agent-id is correct and agent is active
Network errors	Verify API endpoints are accessible and credentials are valid

🐛 Troubleshooting

Debug Mode

Use the --verbose flag to see detailed API request/response information:

tf --verbose send "test message" --agent-id "agent-123"
tf search "test" --verbose

Configuration Issues

1. Check current configuration:

   tf config-show
   

2. Verify credentials:

   tf configure --api-key YOUR_KEY --workspace-id YOUR_WORKSPACE
   

3. Test connection:

   tf chats --verbose
   

📖 More Help

• Command-specific help: tf COMMAND --help
• Detailed examples: tf help-guide
• Verbose debugging: Use --verbose flag
• JSON output: Use --output json for machine-readable output
• Configuration priority: Environment variables → ~/.toothfairy/config.yml → CLI args

✨ Pro Tips

💾 Save time: Configure once with tf configure, then just use tf send and tf search

🔍 Debug issues: Use --verbose to see full API responses and troubleshoot

📝 Scripting: Use --output json and tools like jq to parse responses

⚡ Quick tests: Only --agent-id is required for send, only query for search

🎯 Better search: Use --status, --topics, and --document-id for targeted results

🔧 Multiple environments: Use different config files with --config flag

📄 License

MIT License - see LICENSE file for details.

🔗 Links

• Repository: ToothFairy CLI on Gitea
• Python Package: toothfairy-cli on PyPI
• Node.js Package: @toothfairyai/cli on npm
• Issues: Report bugs and feature requests