Build Telnyx apps faster with AI agent skills

Developers spend 35% of their time reading documentation. Another 20% goes to debugging code they didn't write. When you're integrating a new API, that ratio gets worse: you're constantly switching between docs, code samples, and your editor, trying to piece together the right authentication flow or webhook handler.

AI coding assistants changed this. According to the 2025 Stack Overflow Developer Survey, 84% of developers now use AI tools in their workflow, and 69% report increased productivity when using AI agents. The AI code assistant market hit $3.9 billion in 2025 and is growing at 24% annually.

But AI assistants still have a knowledge problem. They can write generic code, but they don't know the specifics of your APIs. They search the web for code examples but sometimes find outdated or incorrect ones, leading them to hallucinate method names, invent authentication patterns, and produce code that looks right but fails at runtime.

That's why we built Telnyx Agent Skills: plugins that give AI coding assistants deep, accurate knowledge of every Telnyx SDK. Install them once, then describe what you want to build. Your AI writes production code that actually works.

What are agent skills?

Agent skills are plugins for AI coding assistants like Claude Code, Cursor, Windsurf, and other compatible agents. They teach the AI how to correctly use the Telnyx SDKs.

These are official skills for AI coding agents to integrate Telnyx APIs using the native SDKs. They follow the Agent Skills specification and are organized by product and language. Each skill teaches an AI agent how to use Telnyx SDKs correctly with code examples generated from the official OpenAPI specifications.

When you install Telnyx Agent Skills, your AI gains access to:

• Full SDK reference for your language (Python, JavaScript, Go, Java, or Ruby)
• Code examples for all 35 Telnyx products
• Authentication setup and error handling patterns
• Webhook configuration examples
• Best practices for production deployments

Instead of reading docs yourself, you describe what you want. The AI reads the docs and writes production-ready code with proper error handling, retries, and validation.

Quick start: install in 30 seconds

Claude Code

First, add the Telnyx skills marketplace:

/plugin marketplace add team-telnyx/telnyx-ext-agent-skills

Then install the plugin for your language. Each plugin includes all 35 Telnyx products (messaging, voice, numbers, IoT, AI, and more):

Python:

/plugin install telnyx-python@telnyx-agent-skills

JavaScript / Node.js:

/plugin install telnyx-javascript@telnyx-agent-skills

Go:

/plugin install telnyx-go@telnyx-agent-skills

Java:

/plugin install telnyx-java@telnyx-agent-skills

Ruby:

/plugin install telnyx-ruby@telnyx-agent-skills

WebRTC Client SDKs (all platforms):

/plugin install telnyx-webrtc-client@telnyx-agent-skills

Cursor

Add Telnyx skills as project context by referencing the raw SKILL.md files from GitHub:

1. Open Cursor Settings > Rules > Project Rules
2. Create a rule file (e.g., .cursor/rules/telnyx.mdc) and paste the contents of the desired SKILL.md file
3. Alternatively, download the SKILL.md file into your project and add it via @Files in the chat

Windsurf

Add skills to Windsurf via project rules:

1. Create a .windsurfrules file in your project root
2. Paste the contents of the desired SKILL.md file(s) into it

Now try a prompt like: "Create a function that sends an SMS notification when an order ships."

Claude will generate complete, working code that authenticates with your Telnyx API key, validates phone number formats, handles API errors gracefully, and includes retry logic for transient failures.

What you can build

The skills cover all 35 Telnyx products. Here are some prompts to try:

Messaging

• "Send an SMS to +15551234567 saying the order has shipped"
• "Create a webhook server to receive inbound SMS messages"
• "Build a two-way SMS chatbot that responds to customer questions"
• "Send an MMS with an image attachment"

Voice

• "Make an outbound call and play an audio file when they answer"
• "Build an IVR that routes callers based on keypad input"
• "Set up call recording with automatic transcription using Deepgram STT"
• "Create a conference call bridge for up to 20 participants"

AI assistants

• "Create a Voice AI assistant for appointment scheduling"
• "Build an AI agent that handles customer support calls"
• "Configure real-time transcription for a live call"
• "Generate speech from text using Telnyx TTS"

Phone numbers

• "Search for toll-free numbers in the US"
• "Buy a phone number with area code 312 and configure it for SMS"
• "Start a number porting request for my existing number"
• "List all my phone numbers and their current configurations"

Example: building an order notification system

Let's walk through a real example. Say you want to send SMS notifications when orders ship from your e-commerce platform.

Your prompt:

"Create a Python function that sends an SMS notification when an order ships. It should take the customer's phone number, order ID, and tracking URL as parameters. Include proper error handling and logging."

What Claude generates:

import telnyx
import logging
from typing import Optional

logger = logging.getLogger(__name__)

def send_shipping_notification(
    phone_number: str,
    order_id: str,
    tracking_url: str,
    from_number: str = "+18005551234"
) -> Optional[str]:
    """
    Send an SMS notification when an order ships.
    
    Returns the message SID on success, None on failure.
    """
    try:
        message = telnyx.Message.create(
            from_=from_number,
            to=phone_number,
            text=f"Your order #{order_id} has shipped! "
                 f"Track it here: {tracking_url}"
        )
        logger.info(f"Sent shipping notification for order {order_id}")
        return message.id
        
    except telnyx.error.InvalidRequestError as e:
        logger.error(f"Invalid request: {e}")
        return None
    except telnyx.error.AuthenticationError as e:
        logger.error(f"Authentication failed: {e}")
        return None
    except telnyx.error.APIError as e:
        logger.error(f"API error: {e}")
        return None

Notice how Claude:

• Uses the correct Telnyx SDK import and method names
• Includes type hints for clarity
• Handles specific Telnyx error types (not generic exceptions)
• Adds logging for debugging in production
• Returns the message ID for tracking

This is production code, not a tutorial snippet. You can deploy it immediately.

Agent skills vs MCP server

We also offer a Telnyx MCP Server for AI assistants like Claude Desktop. Here's how they differ:

Use Agent Skills when you're building an application and need source code. Use the MCP Server when you want to execute Telnyx operations directly from Claude Desktop without writing code.

Why this matters for developer productivity

The numbers tell the story:

• 69% of developers using AI agents report increased productivity (Stack Overflow 2025)
• 70% say agents reduced time spent on specific development tasks
• 20-40% increase in individual developer output with AI coding assistants
• 51% of professional developers now use AI tools daily

But these gains only happen when the AI has accurate, up-to-date information about your APIs. Generic AI assistants hallucinate SDK methods, invent authentication flows, and produce code that fails at runtime. Agent skills solve this by giving the AI direct access to verified documentation.

Supported products

Telnyx Agent Skills cover the full platform:

• Messaging: SMS, MMS, WhatsApp, short codes, 10DLC
• Voice: Call Control, TeXML, SIP trunking, conferencing
• Numbers: Search, purchase, porting, regulatory compliance
• AI: Voice AI assistants, STT, TTS, embeddings
• Verify: SMS and voice OTP, fraud prevention
• Fax: Programmable fax API
• IoT: SIM management, data plans, connectivity
• Networking: Private connectivity, cloud on-ramps
• Storage: S3-compatible cloud storage
• WebRTC: Browser-based calling

Get started

Agent Skills are open source and free to use. The GitHub repository includes skills for Python, JavaScript, Go, Java, and Ruby.

Install them now and start building with natural language.