Apidance SDK

A Python SDK for interacting with the Apidance API (https://apidance.pro/).

Installation

pip install apidance

Configuration

Create a .env file in your project root with your API credentials:

APIDANCE_API_KEY=your_api_key_here
X_AUTH_TOKEN=your_x_auth_token_here  # Required for reply/like actions

Get your API key from https://apidance.pro

You can find your X_AUTH_TOKEN in your browser cookies when logged into x.com:

1. Open x.com and log in
2. Open browser developer tools (F12 or right-click -> Inspect)
3. Go to Application/Storage -> Cookies -> x.com
4. Find and copy the value of auth_token

Or provide the credentials directly when initializing the client:

client = TwitterClient(
    api_key="your_api_key_here",
    auth_token="your_auth_token_here"  # Required for reply/like actions
)

Usage

Check out the examples

from apidance import TwitterClient

# Initialize the client
client = TwitterClient()

# Search posts (one cursor-paginated page)
response = client.search(query="python", limit=20)
for tweet in response.data:
    print(tweet.author.username, tweet.text, tweet.metrics.like_count)

# Fetch the next page, when available
if response.meta.next_cursor:
    next_page = client.search(query="python", cursor=response.meta.next_cursor)

# Get user information
user = client.get_user_by_screen_name("example")

users = client.get_following(user_id=user["id"])

# Get tweets from a list
list_tweets = client.get_list_latest_tweets(
    list_id="your_list_id",
)

# Reply to a tweet
client.create_tweet(
    text="Your reply text",
    reply_to_tweet_id="tweet_id_to_reply_to",
)

# Like a tweet
client.favorite_tweet(tweet_id="tweet_id_to_like")

MCP server

FillSet config file

{
    "mcpServers": {
        "apidance": {
            "command": "/path/to/uv",
            "args": [
                "--directory",
                "/path/to/apidance-sdk",
                "run",
                "mcp_server.py"
            ]
        }
    }
}

Features

• Search Twitter timeline with various filters (Latest, Top, People, Photos, Videos)
• Get detailed user information by screen name
• Fetch tweets from Twitter lists
• Search Following
• Reply to tweets (requires auth_token)
• Like tweets (requires auth_token)
• Create tweets and reply to existing tweets
• Create note tweets (long rich text tweet, requires Premium+)
• Host as mcp server

Search response (v2)

TwitterClient.search() and the MCP search_tweets tool return a structured response with data and meta. Each post includes its author, ISO 8601 timestamp, metrics, URLs, mentions, media, reply/quote/repost references, and language/source fields. meta.next_cursor is the value to supply to the next search request.

Models

The SDK provides two main data models:

• Tweet: Represents a Twitter post with all its metadata
• User: Contains detailed user information including profile data, stats, and verification status

License

MIT License