Webhooks designed for agents. Create inboxes, wait for events, list and extract them.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Bearer <inbox read token>. Needed for wait, list, get, extract and delete; create_inbox works without it and returns the token.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Create a webhook inbox. Returns its ingest URL and a read token (shown once).
This tool takes no arguments.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http ahook https://ahook.app/mcp
Headers
Authorization (optional, not in the config above): Bearer <inbox read token>. Needed for wait, list, get, extract and delete; create_inbox works without it and returns the token.Create a webhook inbox. Returns its ingest URL and a read token (shown once).
No arguments.
{
"type": "object",
"properties": {}
}Delete an inbox and its events. Auth: Authorization Bearer read token header.
| Argument | Type | Description |
|---|---|---|
| inbox_id* | string |
{
"type": "object",
"required": [
"inbox_id"
],
"properties": {
"inbox_id": {
"type": "string"
}
}
}Pull a JSON path out of an event body. Auth: Authorization Bearer read token header.
| Argument | Type | Description |
|---|---|---|
| event_id* | integer | |
| inbox_id* | string | |
| json_path* | string |
{
"type": "object",
"required": [
"inbox_id",
"event_id",
"json_path"
],
"properties": {
"event_id": {
"type": "integer"
},
"inbox_id": {
"type": "string"
},
"json_path": {
"type": "string"
}
}
}Get one event by id. Auth: Authorization Bearer read token header.
| Argument | Type | Description |
|---|---|---|
| event_id* | integer | |
| inbox_id* | string |
{
"type": "object",
"required": [
"inbox_id",
"event_id"
],
"properties": {
"event_id": {
"type": "integer"
},
"inbox_id": {
"type": "string"
}
}
}List events in an inbox. Auth: Authorization Bearer read token header.
| Argument | Type | Description |
|---|---|---|
| after | integer | |
| limit | integer | |
| inbox_id* | string |
{
"type": "object",
"required": [
"inbox_id"
],
"properties": {
"after": {
"type": "integer"
},
"limit": {
"type": "integer"
},
"inbox_id": {
"type": "string"
}
}
}Wait for the next event in an inbox after a cursor, up to 30s. Auth: Authorization Bearer read token header.
| Argument | Type | Description |
|---|---|---|
| after | integer | |
| inbox_id* | string | |
| timeout_ms | integer | |
| body_contains | string |
{
"type": "object",
"required": [
"inbox_id"
],
"properties": {
"after": {
"type": "integer"
},
"inbox_id": {
"type": "string"
},
"timeout_ms": {
"type": "integer"
},
"body_contains": {
"type": "string"
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 5, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.