Hi n8n team,
I’m hitting what looks like a platform-level failure on the Airtable node’s resource locator in a clean workflow. Details below.
Environment
- n8n Cloud, version 1.123.69
- Instance: [your instance name]
- Node: Airtable (n8n-nodes-base.airtable)
- Credential type: Airtable Personal Access Token API
- Browser: [browser + version]
- Date observed: August 9, 2026, ~10:45 AM (UTC-5)
Problem
In a new workflow, an Airtable node cannot load any dynamic data from the Airtable API through the editor UI.
-
With Base and Table set to “By ID” and Mapping Column Mode set to “Map Each Column Manually”, the node shows:
No columns found in Airtable. Retry. Clicking Retry does not change the result. -
Switching Base to “From list” produces:
Could not load list — No bridge acquired for this context. Call acquire() first.
What I have already ruled out
- The original credential was returning HTTP 401 (
AUTHENTICATION_REQUIRED) from Airtable. I created a brand-new credential of type “Airtable Personal Access Token API” with a freshly issued PAT. - The PAT has the scopes required for this operation, including
schema.bases:read,data.records:read, anddata.records:write, and is explicitly granted access to the target base. - I did a hard reload of the editor tab and reopened the node. The bridge error persists.
- I tested the same credential from an HTTP Request node (v4.3) against
https://api.airtable.com/v0/meta/basesusing Predefined Credential Type → Airtable API, to isolate credential validity from the resource locator path.
Why I think this is not a credential issue
No bridge acquired for this context. Call acquire() first. is not an Airtable API error message — Airtable returns structured JSON errors (401 AUTHENTICATION_REQUIRED, 403 NOT_AUTHORIZED, etc.). This string appears to originate from an internal n8n layer, and the request seems to fail before it ever reaches Airtable.
The same error string is being reported by other Cloud users over the past several weeks, though on a different surface (MCP Server Trigger tool calls rather than the resource locator). That may point to a shared internal component:
- MCP bridge acquisition erro
- Title: MCP Server Trigger (v2) — every tool call fails with "No bridge acquired for this context. Call acquire() first."
- "No bridge acquired for this context. Call acquire() first." Claude - Workday MCP (via n8n)
Questions
- Is this a known issue on Cloud 1.123.x, and is a fix or rollback available?
- Is there a supported workaround that avoids the resource locator entirely while this is open?
- Can you confirm whether the “bridge” failure affects only editor-time dynamic loading (dropdowns, column mapping) or also production workflow executions? This matters for whether I can ship the workflow with hardcoded IDs.