Read AI
Read AI meeting reports: list meetings, fetch one with its summary, action items, transcript and metrics, and read a live meeting's transcript.
About
Read AI ships in the w6w first-party pack. It declares 4 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
- App id
io.w6w.read-ai- Version
- 0.1.0
- Author
- w6w
- Licence
- MIT
- Categories
- AI & Machine Learning · Productivity · Video & Streaming
Overview
Read AI is an AI meeting assistant that joins Zoom, Google Meet and Microsoft Teams calls and produces a report for each one: a summary, chapter summaries, action items, key questions, topics, a transcript and engagement metrics.
This app reads those reports over Read AI’s public REST API (open beta): list meetings newest-first with a start-time window, page through them by cursor, fetch a single meeting with any of the eight expandable fields (summary, chapter summaries, action items, key questions, topics, transcript, metrics, recording download), and pull the live transcript of a meeting still in progress.
The API is read-only and authenticates with OAuth 2.1 (authorization code with refresh tokens) against a client you register yourself through Read AI’s dynamic client registration. Access tokens last ten minutes and refresh tokens rotate on every use, so the connection must be refreshed by the host.
Build with Read AI
Three routes to the same 4 actions. The Workflow tab is generated from Read AI'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.
Get Live Meeting
get-live-meeting Fetch a meeting in progress with its live transcript and chapter summaries. Available only when the live dashboard was open during the meeting.
Get Meeting
get-meeting Fetch one Read AI meeting by ID, optionally with its summary, chapter summaries, action items, key questions, topics, transcript, metrics or recording download.
List All Meetings
list-all-meetings Page through Read AI meetings newest-first, following the cursor, up to a maximum count (10 meetings per API call; rate limit 100 calls/minute).
List Meetings
list-meetings List Read AI meetings newest-first, optionally within a start-time window. One page of at most 10; pass the last meeting's id as the cursor for the next page.
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 Read AI's real ids.
{
"manifestVersion": "2",
"name": "read-ai-example",
"steps": [
{
"id": "get-live-meeting",
"uses": {
"app": "io.w6w.read-ai",
"action": "get-live-meeting",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"meetingId": "<meetingId>"
}
}
]
}Here are some of the things you can do
-
Get Live Meeting
readget-live-meeting -
Get Meeting
readget-meeting -
List All Meetings
readlist-all-meetings -
List Meetings
readlist-meetings
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 Read AI, 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 Read AI. 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: "get-live-meeting",
payload: {
meetingId: "<value>",
// startTimeMsGt: "<value>",
// startTimeMsGte: "<value>",
// expand: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value);npm install -g @w6w/cliw6w run conn_YOUR_CONNECTION_ID --action get-live-meeting --payload '{"meetingId":"<value>"}'Give an AI agent Read AI — without giving it Read AI'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.read-ai#get-live-meeting",
"input": {
"meetingId": "<meetingId>"
}
}
}
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 Read AI connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
What the agent gets
Credentials it can't read
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.
A tool surface scoped to the caller
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.
A durable workflow in one call
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.
Health-aware discovery
Read AI'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. Read AI itself is MIT, and the runtime that executes it is source-available (FSL).
Health checks
Read AI declares its own checks, so its health is a property of the app rather than something the host guesses at.
Read AI platform status
Component status from status.read.ai. The verdict is the API component; meeting bots, sign-in and the apps are reported in the detail but cannot push the verdict past degraded.