Chats, messages, tasks, files, invite links and contacts on Chatwork, Japan's business team chat platform.
Chatwork ships in the w6w first-party pack. It declares 32 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
io.w6w.chatworkChatwork is a business team chat platform, widely used in Japan, that bundles messaging, tasks and file sharing into every chat room. This app covers the full documented surface of the Chatwork API v2: list and manage chats (create, update, leave or delete), read and post messages (including marking them read/unread and editing or deleting your own), create and track tasks assigned within a chat, list and upload files, manage a chat’s invite link, and read or respond to contact requests.
A few behaviors are worth knowing before building a workflow on top of this app: an empty list answers HTTP 204 rather than an empty array (handled transparently by every list action here); the message-list action is a stateful per-connection cursor that returns only what’s new since the last read unless you force a full fetch; and creating a task assigns it to every account listed, returning one task ID per assignee.
Authentication is a personal API token (issued from Chatwork’s own Service Integrations settings, and subject to your organization administrator’s approval). Health checks track the token’s live rate-limit headroom; Chatwork publishes no machine-readable status page, which is declared rather than guessed at.
Three routes to the same 32 actions. The Workflow tab is generated from Chatwork'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.
incoming-request-approve Approve a pending contact request, adding the requester as a contact.
incoming-request-list List pending contact requests sent to the connected account.
my-status-get Fetch unread-message, mention and my-task counters across all of your chats.
room-delete Leave a group chat, or permanently delete it (Administrator role required).
room-file-upload Upload UTF-8 text content to a chat as a new file (vendor limit: 5 MB).
room-link-create Create a chat's invite link. Fails if one already exists — use Update instead.
room-member-update Replace a chat's entire member list. Any current member not included in one of the three role lists is removed from the chat.
room-message-list List a chat's messages. With Force off, returns only messages posted since this connection's last read of this chat (see the action's own notes for what that means).
room-message-mark-read Mark a chat's messages read, up through the given message ID (or the whole chat).
room-message-mark-unread Mark a chat's messages unread, from the given message ID onward.
room-task-create Add a task to a chat. Creates one task per assignee in To — the response lists one task ID per account.
room-update Update a chat's name, description and/or icon. Requires the Administrator role.
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 Chatwork's real ids.
{
"manifestVersion": "2",
"name": "chatwork-example",
"steps": [
{
"id": "incoming-request-approve",
"uses": {
"app": "io.w6w.chatwork",
"action": "incoming-request-approve",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"requestId": "<requestId>"
}
}
]
}incoming-request-approve contact-list incoming-request-list incoming-request-reject me-get +27 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 Chatwork, 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 Chatwork. 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: "incoming-request-approve",
payload: {
requestId: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action incoming-request-approve --payload '{"requestId":"<value>"}' Give an AI agent Chatwork — without giving it Chatwork'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.chatwork#incoming-request-approve",
"input": {
"requestId": "<requestId>"
}
}
}
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 Chatwork 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.
Chatwork'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. Chatwork itself is MIT, and the runtime that executes it is source-available (FSL).
Chatwork declares its own checks, so its health is a property of the app rather than something the host guesses at.
No machine-readable status page or incident feed was found for Chatwork as of 2026-08-29 (checked: status.chatwork.com, chatworkstatus.com and info.chatwork.com, none of which resolve; chatwork.statuspage.io, which is an unclaimed Statuspage instance; and the developer portal's own guide pages, which name none).
Requests remaining in the current window, read from the X-RateLimit-* headers on GET /me.