Flatfile for AI agents
Flatfile is a data exchange platform for collecting, validating, and transforming structured data across collaborative import workflows, used by product teams onboarding customer data. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.
What agents do in Flatfile.
Check an import job's status
Get a job's status to confirm whether a customer's uploaded file has finished validating.
Pull records from a sheet
List records from a sheet to review data before it is pushed into a production system.
Confirm before deleting records
Require approval before deleting records, since it permanently removes rows from the import.
12 Flatfile actions, graded by risk.
Every Flatfile action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.
read
9Look things up. Allowed by default.
- flatfile.get_jobGet Job
- flatfile.get_sheetGet Sheet
- flatfile.get_spaceGet Space
- flatfile.list_jobsList Jobs
- flatfile.list_filesList Files
- flatfile.list_spacesList Spaces
- flatfile.list_recordsList Records
- flatfile.list_workbooksList Workbooks
- flatfile.list_environmentsList Environments
write
2Create and change things. Allow, or ask the user first.
- flatfile.update_recordsUpdate Records
- flatfile.insert_recordsInsert Records
destructive
1Delete, cancel or archive. Ask first, or deny outright.
- flatfile.delete_recordsDelete Records
Flatfile in three steps.
- 01Your users connect FlatfileThey add their Flatfile api key on Arc0 Connect, under your brand. It goes straight into the vault.
- 02You set the rulesReads run, writes like “update Records” can wait for the user, and “delete Records” can be denied outright.
- 03Any agent can actYour agent calls Flatfile through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
await arc0.policies.set('flatfile', { read: 'allow', write: 'ask', // update_records destructive: 'deny', // delete_records }) # Claude Code: the same connection, one URL $ claude mcp add --transport http arc0 \ https://mcp.arc0.ai/u/u_8f2
How Flatfile connects.
Users add their Flatfile api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
The same Flatfile connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →
- AUTH
- API key
- CREDENTIALS
- Per-tenant encrypted vault
- MODEL SEES
- Results only, never credentials
- AUDIT LOG
- Every call, on every plan
Use Flatfile from any agent.
ClaudeChatGPTCursor
Codex
VS CodeOpenAI Agents SDKClaude Agent SDK
Vercel AI SDK
Mastra
LangGraph
Flatfile and Arc0, answered.
Can I use Flatfile with Claude, ChatGPT or Cursor?
Yes. Connect Flatfile to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Flatfile actions you allow.
How do users connect Flatfile?
Users add their Flatfile api key on Arc0 Connect. It is encrypted in the vault, never shown to the model, and each user can rotate or revoke it at any time.
Which Flatfile actions can my agent take?
12 in total: 9 read, 2 write and 1 destructive, such as “update Records”. Your policies decide which of them each agent may call.
Can I stop my agent from deleting things in Flatfile?
Yes. Actions like “delete Records” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.
Can my own backend call Flatfile too?
Yes. The same Flatfile connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.
Plug Flatfile into your agent.
Your users connect Flatfile once, under your brand. Your agent gets 12 actions behind your policies, with every call on the record.
Free to build · MCP + REST · Audit log on every plan