| name | callme |
| description | Message the user on Telegram when you need input, want to report progress, or need real-time discussion. Use for completed tasks, blocking questions, or milestone celebrations. |
| allowed-tools | ["mcp__callme__initiate_call","mcp__callme__continue_call","mcp__callme__speak_to_user","mcp__callme__end_call"] |
Telegram Message Input Skill
Description
Message the user on Telegram for real-time text conversations. Use this when you need input, want to report on completed work, or need to discuss next steps.
When to Use This Skill
Use when:
- You've completed a significant task and want to report status and ask what's next
- You need real-time user input for complex decisions
- A question requires back-and-forth discussion to fully understand
- You're blocked and need urgent clarification to proceed
- You want to celebrate a milestone or walk the user through completed work
- A task will take a long time and the user should be notified when it's done
Do NOT use for:
- Information the user has already provided
- Every minor step — batch updates for significant milestones
Tools
initiate_call
Start a Telegram conversation with the user.
Parameters:
message (string): What you want to say. Be natural and conversational.
Returns:
- Conversation ID and the user's text response
continue_call
Continue an active conversation with a follow-up message.
Parameters:
call_id (string): The conversation ID from initiate_call
message (string): Your follow-up message
Returns:
speak_to_user
Send a message without waiting for a response. Use this to acknowledge requests or provide status updates before starting time-consuming operations.
Parameters:
call_id (string): The conversation ID from initiate_call
message (string): What to send to the user
Returns:
- Confirmation that the message was sent
When to use:
- Acknowledge a request before starting a long operation (e.g., "Let me search for that...")
- Provide status updates during multi-step tasks
- Keep the conversation flowing naturally so the user knows you're working
end_call
End an active conversation with a closing message.
Parameters:
call_id (string): The conversation ID from initiate_call
message (string): Your closing message (say goodbye!)
Returns:
- Conversation duration in seconds
Example Usage
Simple conversation:
1. initiate_call: "Hey! I finished the auth system. Should I move on to the API endpoints?"
2. User responds: "Yes, go ahead"
3. end_call: "Perfect! I'll start on the API endpoints. Talk soon!"
Multi-turn conversation:
1. initiate_call: "I'm working on payments. Should I use Stripe or PayPal?"
2. User: "Use Stripe"
3. continue_call: "Got it. Do you want the full checkout flow or just a simple button?"
4. User: "Full checkout flow"
5. end_call: "Awesome, I'll build the full Stripe checkout. I'll let you know when it's ready!"
Using speak_to_user for long operations:
1. initiate_call: "Hey! I finished the database migration. What should I work on next?"
2. User: "Can you look up the latest API documentation for Stripe?"
3. speak_to_user: "Sure! Let me search for that. Give me a moment..."
4. [Perform web search and gather information]
5. continue_call: "I found the latest Stripe API docs. They released v2024.1 with new payment methods..."
6. User: "Great, implement that"
7. end_call: "Perfect! I'll implement the new payment methods. Talk soon!"
Best Practices
- Be conversational — Talk naturally, like a real conversation
- Provide context — Explain what you've done before asking questions
- Offer clear options — Make decisions easy with specific choices
- Use speak_to_user for acknowledgments — Before time-consuming operations (searches, file reads, etc.), use
speak_to_user to acknowledge the request so the user isn't left wondering what's happening
- Always end gracefully — Say goodbye and state what you'll do next