FlurryPORT — Claude Code MCP server: webhooks and agent rooms
Catches webhooks from Stripe, GitHub and Shopify, shows what landed, and replays it into your app with the signature still valid.
claude mcp add --transport http webhooks-and-rooms https://mcp.flurryport.io/mcpWhat FlurryPORT does
FlurryPORT hands you a stable capture URL. Point a provider at it, watch the events pile up, then send any of them into your service again as often as you like. Original method, headers and body travel along, so signature checks pass. No tunnel involved.
The MCP toolset covers one loop: make an endpoint, wait for a capture, read it, reshape it with JSONata, replay it to a target. An anonymous session carries 250 captures and can be claimed later. Have an account already? flurryport login stores the token and the server picks it up in authenticated mode. Sixty tools exist; twenty of them answer before the session is claimed.
Replays reach loopback addresses only unless you pass --allow-lan to the mcp command — needed when your dev server sits on another machine. Hosted chat clients cannot start a local process, so stdio is out for them; that is what the seat server is for.
What it helps with
- receive provider webhooks with no ngrok tunnel
- read the exact headers and body a provider sent
- replay an event into the app without breaking its signature
- generate a test event shaped like the provider's payload
- prove the handler with a local echo server before writing code
Example prompts for the agent
- Start the flurryport MCP server and create an endpoint for Stripe webhooks
- Show the latest captured events and replay the newest one into my local server
- Run an echo server on port 3000, send it the last webhook and print the headers and body
How to install FlurryPORT
1Platform CLI: remote server
claude mcp add --transport http webhooks-and-rooms https://mcp.flurryport.io/mcp2Platform CLI
claude mcp add flurryport -- npx -y flurryport1Platform CLI: remote server
codex mcp add flurryport --url https://mcp.flurryport.io/mcp2Platform CLI
codex mcp add flurryport -- npx -y flurryportConfig file
{
"mcpServers": {
"flurryport": {
"url": "https://mcp.flurryport.io/mcp"
}
}
}{
"mcpServers": {
"flurryport": {
"command": "npx",
"args": [
"-y",
"flurryport"
]
}
}
}Config file
{
"servers": {
"flurryport": {
"type": "http",
"url": "https://mcp.flurryport.io/mcp"
}
}
}{
"servers": {
"flurryport": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"flurryport"
]
}
}
}1Platform CLI: remote server
gemini mcp add -t http flurryport https://mcp.flurryport.io/mcp2Platform CLI
gemini mcp add flurryport npx -y flurryportStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about FlurryPORT
How do I install FlurryPORT in Claude Code?
Use the hosted server: claude mcp add --transport http flurryport https://mcp.flurryport.io/mcp. Or register the local one via npx: claude mcp add flurryport -- npx -y flurryport.
What does FlurryPORT do?
It captures webhooks on a fixed URL, lets you inspect them, and replays them into your app with the same headers, body and signature. It also runs signed rooms where several agents write to one shared record.
Why don't webhooks reach my dev server?
Forwarding is limited to loopback addresses by default. For a dev server on another machine in your network, add --allow-lan to the mcp command.
