Airtable
Spreadsheet-database hybrid for ops.
Verdict
Common use cases
- Log customer feedback into a CRM base
- Bulk-create project tasks from meeting notes
- Add comments to bug tracker records
- Spin up a new base for a product launch
- Delete test records after a demo
Integration
- Vendor
- Airtable
- Category
- productivity
- Auth
- OAUTH2
- Tools
- 17
- Composio slug
airtable
Tools
- Create a record
Creates a new record in a specified airtable table; field values must conform to the table's column types.
- Create base
Creates a new airtable base with specified tables and fields within a workspace; ensure field options are valid for their type.
- Create Comment
Creates a new comment on a specific record within an airtable base and table.
- Create Field
Creates a new field within a specified table in an airtable base.
- Create multiple records
Creates multiple new records in a specified airtable table.
- Create table
Creates a new table within a specified existing airtable base, allowing definition of its name, description, and field structure.
- Delete Commentdestructive
Deletes an existing comment from a specified record in an airtable table.
- Delete multiple recordsdestructive
Deletes up to 10 specified records from a table within an airtable base.
- Delete Recorddestructive
Permanently deletes a specific record from an existing table within an existing airtable base.
- Get Base Schema
Retrieves the detailed schema for a specified airtable base, including its tables, fields, field types, and configurations, using the `baseid`.
- Get Record
Retrieves a specific record from a table within an airtable base.
- Get user information
Retrieves information, such as id and permission scopes, for the currently authenticated airtable user from the `/meta/whoami` endpoint.
- List bases
Retrieves all airtable bases accessible to the authenticated user, which may include an 'offset' for pagination.
- List Comments
Retrieves all comments for a specific record in an airtable table, requiring existing `baseid`, `tableidorname`, and `recordid`.
- List records
Retrieves records from an airtable table, with options for filtering, sorting, pagination, and specifying returned fields.
- Update multiple records
Updates multiple existing records in a specified airtable table; these updates are not performed atomically.
- Update record
Modifies specified fields of an existing record in an airtable base and table; the base, table, and record must exist.
Setup
Setup guide
- 11. Open any Space in Switchy and click the integrations menu in the sidebar. 2. Find Airtable in the productivity category and click Connect. 3. You'll land on Airtable's OAuth consent screen — sign in with the account that owns the bases you want to access. 4. Grant the requested scopes: data.records:read, data.records:write, schema.bases:read, schema.bases:write, and comments:write. 5. Airtable redirects you back to Switchy with a green confirmation banner. 6. Return to your Space and type '@Airtable list my bases' to confirm the connection works. 7. To create a record, @mention Airtable and specify the base ID, table name, and field values in your prompt. 8. Check the response for the new record ID and a link back to Airtable's web UI.
What teammates see: by default, memories from Airtable are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Create Feedback Record
@Airtable create a record in base appXYZ123, table 'Customer Feedback'. Set 'Quote' to 'Love the new dashboard', 'Source' to 'Email', and 'Sentiment' to 'Positive'.Open in a Space →
Bulk Add Tasks
@Airtable create multiple records in base appABC456, table 'Tasks'. Add five records with 'Name' set to 'Task 1' through 'Task 5' and 'Status' set to 'Not Started'.Open in a Space →
Comment on Bug Report
@Airtable create a comment on record recDEF789 in base appGHI012, table 'Bugs'. Comment text: 'Reproduced on staging, escalating to engineering'.Open in a Space →
Build Launch Tracker Base
@Airtable create a base named 'Q2 Product Launch' in workspace wspJKL345. Add a table called 'Milestones' with fields: 'Name' (single line text), 'Owner' (single line text), 'Due Date' (date).Open in a Space →
Clean Up Test Data
@Airtable delete records recMNO678, recPQR901, recSTU234 from base appVWX567, table 'Staging Data'.Open in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@airtable show me all records in the Product Roadmap base where status is 'In Progress' and priority is high
Found 7 records matching your criteria in the Product Roadmap base: 1. Mobile app redesign (Priority: High, Owner: Sarah Chen, Due: Feb 15) 2. API rate limiting improvements (Priority: High, Owner: Marcus Liu, Due: Feb 22) 3. Dashboard performance optimization (Priority: High, Owner: Priya Patel, Due: Mar 1) 4. SSO integration for enterprise (Priority: High, Owner: James Park, Due: Mar 8) 5. Webhook reliability fixes (Priority: High, Owner: Sarah Chen, Due: Mar 15) 6. Analytics export feature (Priority: High, Owner: Marcus Liu, Due: Mar 22) 7. Mobile push notifications (Priority: High, Owner: Priya Patel, Due: Apr 5) All items are currently in progress with assigned owners and target dates within the next 8 weeks.
This example demonstrates read-heavy querying with filters. The MCP requires OAuth2 scope for the specific base being accessed. Response shape and available fields depend entirely on your Airtable schema — this output assumes a typical roadmap structure with Status, Priority, Owner, and Due Date columns.
@airtable create 5 new records in our Customer Feedback table: one for each piece of feedback from today's user interviews, including contact email and sentiment
Successfully created 5 new records in the Customer Feedback table: • Record 1: Contact sarah.m@techcorp.com, Sentiment: Positive, Feedback: "Love the new dashboard layout" • Record 2: Contact james.k@startup.io, Sentiment: Neutral, Feedback: "Export feature works but could be faster" • Record 3: Contact priya.s@enterprise.com, Sentiment: Negative, Feedback: "Mobile app crashes on Android 13" • Record 4: Contact marcus.l@agency.co, Sentiment: Positive, Feedback: "API documentation is excellent" • Record 5: Contact emma.w@nonprofit.org, Sentiment: Neutral, Feedback: "Would like more customization options" All records are now visible in your Airtable base and available for filtering, sorting, and reporting.
This showcases bulk record creation, which is more efficient than individual creates. Field values must match your table's column types exactly (single-line text, email, single-select, etc.). If field types don't align, the operation fails. The MCP can create up to 10 records per request due to Airtable API limits.
@airtable review all comments on the Q1 Planning record and draft a summary of team concerns, then add that summary as a new comment
I reviewed 8 comments on the Q1 Planning record. Key themes: **Resource Constraints** (4 mentions): Team flagged insufficient engineering bandwidth for mobile and API work simultaneously. Sarah and Marcus both noted their teams are at capacity through March. **Timeline Concerns** (3 mentions): Priya questioned whether Feb 15 launch is realistic given current velocity. James suggested pushing enterprise features to Q2. **Dependency Risks** (2 mentions): Emma highlighted that analytics export depends on the unfinished dashboard work. I've added this summary as a new comment on the Q1 Planning record. The team can now reference this consolidated view instead of re-reading individual threads.
This example pairs the MCP's read capability (fetching comments) with AI synthesis, then writes back to Airtable. It demonstrates workflow automation — the AI acts as a meeting notes compiler. Requires comment read/write scope. Be aware that creating comments triggers Airtable notifications to collaborators, so this has side effects beyond data storage.
Use-case deep-dives
When Airtable MCP beats spreadsheets for support ops
A 6-person support team logs customer feedback in Airtable with tags, severity, and linked tickets. During weekly triage, the team lead asks Switchy to pull all high-severity items from the last sprint, create follow-up records for engineering, and post summary comments on each. The MCP's multi-record create and comment tools handle this in one prompt instead of 40 manual clicks. The OAuth2 setup means no API key rotation headaches. This works until you hit 500+ feedback items per week—at that scale, Airtable's rate limits start to bite and you'll want batch export instead. If your triage is under 200 records and you already live in Airtable, this MCP turns standup prep from 20 minutes to 2.
Why small content teams pick Airtable MCP for publishing workflows
A 3-person content team tracks blog posts, social threads, and email campaigns in a single Airtable base with linked records for authors and channels. The marketer asks Switchy to create next month's calendar skeleton—12 blog post records with draft titles, assigned writers, and placeholder publish dates—then add a comment on each flagging SEO keyword gaps. The MCP's create-multiple-records and create-comment tools make this a 30-second ask instead of a 15-minute copy-paste session. The create-field tool also lets you add new columns mid-sprint without leaving the chat. This breaks down if your calendar spans multiple bases or you need complex rollup formulas—Airtable's API doesn't expose those. For single-base publishing workflows under 50 assets per month, this MCP is the fastest path from idea to scheduled post.
When Airtable MCP simplifies recruiter handoffs at small scale
A 10-person startup tracks candidates in Airtable with stages, interview notes, and offer details. The hiring manager asks Switchy to move 5 candidates to 'Final Round', create interview schedule records for each, and post a comment tagging the recruiter with next steps. The MCP's update, create, and comment tools handle the full handoff in one prompt. OAuth2 means the manager's permissions flow through—no shared API key risk. This works cleanly until you're hiring for 10+ roles at once or need custom views per interviewer—at that point, Airtable's interface limits start to show and you'll want a dedicated ATS. For early-stage teams hiring 2-5 people per quarter, this MCP turns pipeline updates from a 10-minute admin task into a single conversational ask.
Frequently asked
What can the Airtable MCP do in Switchy?
The Airtable MCP lets your AI agents create and modify Airtable bases, tables, fields, and records directly from Switchy. Agents can add rows, build new tables with custom field types, post comments on records, and delete entries — all without leaving the conversation. It's useful when your team stores structured data in Airtable and wants AI to update it on the fly.
Do I need admin access to connect Airtable via OAuth?
You need workspace creator or owner permissions to create new bases and tables. For reading and writing records in existing bases, editor access is enough. During OAuth setup, Airtable will show you exactly which workspaces and bases the MCP can touch. If you only have commenter access, the MCP won't be able to create or delete records for you.
Can the Airtable MCP read existing records or just create new ones?
This MCP focuses on writes — creating bases, tables, fields, records, and comments, plus deleting records and comments. It doesn't expose a read or query tool in the current toolset. If your agent needs to search Airtable data before acting, you'll need to fetch that context separately or use Airtable's API directly alongside this MCP.
Why use this instead of Airtable's API or automations?
Airtable's API requires you to write and maintain code; their automations are click-based and limited to predefined triggers. The MCP lets your AI decide when to update Airtable based on conversation context — no scripts, no if-this-then-that rules. It's faster for ad-hoc data entry and structure changes, but less predictable than a scheduled automation.
Who on the team should connect the Airtable MCP?
Whoever owns the Airtable workspace your team uses day-to-day. That person's OAuth token determines which bases and tables the AI can touch. If multiple people need AI-driven Airtable access, each can connect their own Airtable account in Switchy, and the MCP will respect their individual permissions when agents run tools.