Process Street offers a Model Context Protocol (MCP) server, accessible via app integration, which allows AI assistants and other applications to access your workflows, workflow runs, tasks, users, and data sets from outside the Process Street platform.
This server provides a way to interact with your Process Street organization through various AI platforms and tools that support MCP.
With this server, Process Street customers can:
Example requests users could make from outside Process Street using the MCP:
The MCP server supports most Process Street API endpoints as tools.
The MCP server uses your Process Street login credentials to access it, including organizations with SAML SSO enabled.
Admins, Builders and Users can use the MCP server and can only interact with resources they have permissions to.
A compatible MCP client, including:
Note: MCP support in ChatGPT requires a Pro, Team, Business, Enterprise, or Edu plan.
Tip: ChatGPT asks you to confirm before executing any actions that modify data in Process Street (such as creating runs or updating tasks). Read-only actions like viewing workflows or pulling reports run without a confirmation prompt.
When you connect an MCP client for the first time (through Claude, ChatGPT, or Copilot Studio), Process Street displays an authorization screen. This screen shows the name of the application requesting access, the permissions (scopes) it needs, and the redirect URI — so you can confirm you recognize the application before granting access.
Click Approve to grant the application access to your Process Street data, or Cancel to deny the request. Canceling returns you to the connecting application without granting access.
If you’ve previously authorized the same client with the same permissions, the consent screen doesn’t appear again — access is granted automatically using your existing authorization. To require re-authorization, revoke the application’s access first; the consent screen reappears on the next connection attempt.
Process Street’s MCP server works with any application that supports the Model Context Protocol. For clients we don’t offer a one-click connector for, add it as a remote MCP server and authenticate with a bearer token, using a Process Street API key as the token.
You’ll need two things:
https://mcp.process.st/Then add an Authorization: Bearer YOUR_API_KEY header to your client’s MCP configuration. Two common examples follow. Treat your API key like a password.
Add the server to .vscode/mcp.json in your workspace, or open your user configuration with the MCP: Open User Configuration command. Using an input keeps the key out of the file: VS Code prompts for it once and stores it securely.
{
"servers": {
"process-street": {
"type": "http",
"url": "https://mcp.process.st/",
"headers": {
"Authorization": "Bearer ${input:ps-api-key}"
}
}
},
"inputs": [
{
"id": "ps-api-key",
"type": "promptString",
"description": "Process Street API key",
"password": true
}
]
}Add the server to ~/.cursor/mcp.json (available in every project) or .cursor/mcp.json (for a single project), then replace YOUR_API_KEY with your Process Street API key.
{
"mcpServers": {
"process-street": {
"url": "https://mcp.process.st/",
"headers": {
"Authorization": "Bearer YOUR_API_KEY"
}
}
}
}The same pattern, a remote MCP server with an Authorization: Bearer header, works for any other MCP client that supports remote servers.
If you’re using an MCP client that we don’t already support and need a hand, reach out to our support team.
Help us improve this help center.
