Read, create, update, upsert and delete Airtable records; list bases and get base schema.
Airtable ships in the w6w first-party pack. It declares 10 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
io.w6w.airtableAirtable turns a workflow into a full read/write client for any base — creating, updating, upserting and deleting records without hand-rolling the field-type quirks Airtable’s API is known for. Records can be searched with a formula filter, or appended in bulk, or reconciled through the upsert action so a sync doesn’t have to choose between create and update ahead of time.
Base and table structure is available too: listing every base a token can see and pulling a table’s live schema, so a workflow can adapt its own field mapping instead of hardcoding column names that drift when someone edits a view.
It fits any process built around Airtable as a lightweight database — turning a form submission into a new record, keeping a CRM view in sync with an external source, or reading structured data out of a base for another step in a workflow to consume.
Three routes to the same 10 actions. The Workflow tab is generated from Airtable'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.
append-records Append up to 10 records in one request. For >10 records, call this action in a loop.
search-records List or filter records. Returns one page with an `offset` cursor for the next page.
update-record Update a single record. Set `replace: true` to overwrite unspecified fields with null.
upsert-records Create-or-update up to 10 records, matched on the given `fieldsToMergeOn`.
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 Airtable's real ids.
{
"manifestVersion": "2",
"name": "airtable-example",
"steps": [
{
"id": "append-records",
"uses": {
"app": "io.w6w.airtable",
"action": "append-records",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"baseId": "<baseId>",
"table": "<table>",
"records": "<records>"
}
}
]
}append-records create-record get-base-schema get-record get-table +5 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 Airtable, 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 Airtable. 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: "create-record",
payload: {
baseId: "<value>",
table: "<value>",
fields: "<value>",
// typecast: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value); npm install -g @w6w/cli w6w run conn_YOUR_CONNECTION_ID --action create-record --payload '{"baseId":"<value>","table":"<value>","fields":"<value>"}' Give an AI agent Airtable — without giving it Airtable'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.airtable#append-records",
"input": {
"baseId": "<baseId>",
"table": "<table>",
"records": "<records>"
}
}
}
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 Airtable 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.
Airtable'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. Airtable itself is MIT, and the runtime that executes it is source-available (FSL).
Airtable declares its own checks, so its health is a property of the app rather than something the host guesses at.
Atlassian Statuspage rollup for status.airtable.com, with per-component detail. Unauthenticated and unsigned.