Turn incoming emails and documents into structured data with Parseur's AI mailboxes, and manage the templates, webhooks and custom downloads that shape what comes out.
Parseur ships in the w6w first-party pack. It declares 29 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
io.w6w.parseurParseur turns unstructured emails, PDFs, images and spreadsheets into structured JSON. A mailbox (called a “parser” throughout the API) receives documents — forwarded emails, direct file uploads, or raw email/text posted straight to the API — and an AI or template engine extracts the fields you configure into a document record.
This app covers the full lifecycle: create and configure mailboxes, list and inspect the documents they’ve received (including per-document processing logs), submit new documents either as a binary upload or as an email/text payload, and drive documents through their own lifecycle (reprocess, skip, copy to another mailbox). It also manages the templates a mailbox uses to identify layouts, the webhooks that push parsed data out the moment a document finishes, and the custom CSV/XLS downloads Parseur can generate per mailbox or per field.
Authentication is a single account-wide API key, and Parseur’s own health signal is the same call used to prove that key is live: hitting the API root returns a tiny routing map with no way to leak anything, so this app also uses it as its liveness probe rather than any endpoint that touches real mailbox data.
Three routes to the same 29 actions. The Workflow tab is generated from Parseur's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
bootstrap-get Get Parseur's static configuration: enum choices, field length limits, the inbound email domain, and the catalog of pre-built master mailbox templates. This is a public endpoint.
document-upload Upload a binary file (PDF, image, EML, ...) to a mailbox for parsing. Processing happens asynchronously after this call returns.
email-create Submit an email/text payload as a new document, without sending real email. The target mailbox's own address must appear in Recipient, To, Cc, or Bcc — see the action's notes.
export-config-create Create a custom CSV/XLS download configuration on a mailbox.
export-config-list List the custom CSV/XLS download configurations on a mailbox.
mailbox-delete Delete a mailbox by id. Deletion is asynchronous — see the action's notes.
webhook-create Register a webhook. Use Enable Webhook afterwards to attach it to a mailbox.
webhook-disable Detach a webhook from a mailbox without deleting the webhook registration.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Parseur's real ids.
{
"manifestVersion": "2",
"name": "parseur-example",
"steps": [
{
"id": "document-copy",
"uses": {
"app": "io.w6w.parseur",
"action": "document-copy",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"documentId": "<documentId>",
"targetMailboxId": "<targetMailboxId>"
}
}
]
}document-copy bootstrap-get document-get document-list document-log-list +24 more actions available
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Parseur, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Parseur. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "document-copy",
payload: {
documentId: "<value>",
targetMailboxId: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action document-copy --payload '{"documentId":"<value>","targetMailboxId":"<value>"}' Give an AI agent Parseur — without giving it Parseur's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.parseur#document-copy",
"input": {
"documentId": "<documentId>",
"targetMailboxId": "<targetMailboxId>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Parseur connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Parseur's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Parseur itself is MIT, and the runtime that executes it is source-available (FSL).
Parseur declares its own checks, so its health is a property of the app rather than something the host guesses at.