| name | airtable-connector |
| description | Build, operate, or harden a schema-aware Airtable connector for bases, tables, fields, record queries, exports, and confirmed writes. Use when a request explicitly concerns Airtable integration or Airtable record operations; do not use for generic database or spreadsheet work. |
Airtable Connector
Use this skill when Prometheus needs to read, search, create, update, or export Airtable records.
Prometheus Fit
Airtable should be a REST connector with schema-aware tools. Prometheus needs field metadata before it can safely generate writes.
Tool Scope
Start with:
airtable_list_bases
airtable_list_tables
airtable_get_schema
airtable_search_records
airtable_get_record
airtable_create_record
airtable_update_record
airtable_export_table
Rules
- Fetch table schema before writing records.
- Validate field names and field types before create/update.
- Use Airtable pagination and return offsets/cursors.
- Store API credentials only through Connections/vault.
- Creating, updating, deleting, or bulk-changing records requires explicit confirmation.
- Treat formulas, linked records, attachments, and collaborators as typed fields, not strings.
- Return record IDs and Airtable URLs when possible.
Implementation Route
- Use
connector-builder to scaffold a REST connector.
- Add schema and read/search tools before writes.
- Implement field validation in the connector layer.
- Add CSV/JSON export to workspace artifacts for analysis flows.
- Add mocked tests for pagination and field validation.
Acceptance Check
Prometheus can inspect Airtable structure, query records, and propose writes without guessing field shapes.