Official integrations
MCP Server
The PingRep MCP Server connects AI Representatives to Claude Desktop, Cursor, and other MCP-compatible tools. Install it from npm:
After installation, add the server to your MCP client configuration:
The MCP Server exposes tools for:
| Tool | Description |
|---|---|
search_directory | Search the PingRep professional directory |
get_profile | Retrieve a professional profile by slug |
ask_representative | Ask a question to someone's AI Representative |
send_message | Send a chat message to an AI Representative |
MCP documentation
For a complete walkthrough, see the MCP Integration guide.
REST API
The PingRep REST API is the primary integration method. No SDK installation required. Use any HTTP client in any language.
Community libraries
Community contributions welcome
We do not yet have official SDKs for specific languages. If you build a client library for the PingRep API, let us know at support@pingrep.com and we will list it here.
Community-maintained libraries will be listed here as they become available. If you are building one, we recommend:
- Following the API Keys guide for authentication
- Implementing retry logic per the Rate Limits guide
- Supporting both synchronous and streaming response modes for the AI endpoints
- Including TypeScript type definitions if applicable
API client tips
Regardless of language, keep these patterns in mind when building your own client:
- Set a base URL constant. All requests go to
https://api.pingrep.com. - Include auth on every request. Pass your API key or JWT as a Bearer token in the Authorization header.
- Handle rate limits gracefully. Parse the
X-RateLimit-*headers and implement backoff. See Rate Limits. - Parse error responses consistently. The
detailfield can be a string or an array of validation errors. See Error Codes.