Connect AI Agents to any API
Discover, execute, and manage API actions with agents.json. Build custom collections and use our growing official library of agentic APIs.
Giphy
Search and share GIFs
Slack
Send messages and manage channels
Stripe
Process payments securely
Your AI's Gateway to APIs
Discover and execute API actions across thousands of endpoints intelligently
Custom Collections
Start with our collection of agents.json specifications, ready to use with your AI agents.
Create custom collections by adding your own APIs and tools with agents.json.
Tool Search
LLMs struggle to select the right action beyond a few tools.
Let your AI agents query Wildcard to pick tools using natural language.
Tool Selection Accuracy
Model-Agnostic Execution
Wildcard drops in tool definitions catered to your agent's model.
Own your tool calls. Execute actions on your own infrastructure.
agents.json registry
Discoverable and Optimized API Actions for your AI agents.
Resend API
Deliver transactional and marketing emails at scale.
Alpaca Markets API
Trade stocks and crypto with real-time market data.
Rootly API
AI-powered on-call and incident response platform
Slack API
Send messages, manage channels and conversations
Twitter API
Post tweets, fetch user data, and interact with Twitter content.
Stripe API
Manage products, subscriptions, and customers.
HubSpot Contacts API
Create, update, and manage CRM contacts.
Google Sheets API
Read, write, and update spreadsheet data
Giphy API
Search, retrieve, and embed GIFs and animated stickers.
The Odds API
Retrieve real-time sports odds, scores, and betting market data.
Get Started in Minutes
Install our SDK and start building with agents.json right away
1. Install the SDK
pip install agentsjson2. Search for Actions
import requests
flow = requests.get(
"https://api.wild-card.ai/search",
params={
"query": "Create a product called
Premium Access for $100",
"collection_id": "<YOUR_COLLECTION_ID>",
}
)3. Execute Actions
from agentsjson import execute, get_tool
from openai import OpenAI
tools = get_tools(flow)
response = openai.chat.completions.create(
model="gpt-4o",
messages=[{
"role": "user",
"content": "Create a product called
Premium Access for $100"
}],
tools=tools
)
auth = BearerAuth(token="<your-token>")
response = execute(flow, response, auth)Frequently Asked Questions
Common questions about agents.json and how it works