Extracta.ai for AI agents

Extracta.ai automates data extraction from PDFs, images, and text files using AI, without needing prior document training, used by teams processing varied paperwork at speed. 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.

Connect Extracta.ai See the 10 actions ↓

DocumentsAPI keyMCP + RESTextracta.ai

AUDIT LOG · EXTRACTA.AIPOLICY: acme-support

09:41:07 · claude · u_8f2read

extracta_ai.get_credits

Get Credits✓ allowed · 212ms

09:41:08 · claude · u_8f2read

extracta_ai.view_extraction

View Extraction✓ allowed · 164ms

09:41:09 · claude · u_8f2write

extracta_ai.create_extraction

Create Extraction✓ approved · approved by user

09:41:12 · claude · u_8f2destructive

extracta_ai.delete_extraction

Delete Extraction✕ blocked · policy: deny

EVERY EXTRACTA.AI CALL, ON THE RECORD

01 · USE CASES

What agents do in Extracta.ai.

Extract data from a document

Run an extraction on an uploaded document to pull structured fields into another system.

Check a batch's results

Get the results of a batch extraction job once processing on a set of files finishes.

Check credits before a large batch

Read remaining credits before submitting a large batch of documents for extraction.

02 · ACTIONS

10 Extracta.ai actions, graded by risk.

Every Extracta.ai action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

Look things up. Allowed by default.

extracta_ai.get_credits
Get Credits
extracta_ai.get_batch_results
Get Batch Results
extracta_ai.view_extraction
View Extraction
extracta_ai.view_classification
View Classification

write

Create and change things. Allow, or ask the user first.

extracta_ai.create_extraction
Create Extraction
extracta_ai.update_extraction
Update Extraction
extracta_ai.create_classification
Create Classification
extracta_ai.update_classification
Update Classification

destructive

Delete, cancel or archive. Ask first, or deny outright.

extracta_ai.delete_extraction
Delete Extraction
extracta_ai.delete_classification
Delete Classification

03 · HOW IT WORKS

Extracta.ai in three steps.

01Your users connect Extracta.aiThey add their Extracta.ai api key on Arc0 Connect, under your brand. It goes straight into the vault.
02You set the rulesReads run, writes like “create Extraction” can wait for the user, and “delete Extraction” can be denied outright.
03Any agent can actYour agent calls Extracta.ai through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.

POLICY.TS

await arc0.policies.set('extracta_ai', {
  read: 'allow',
  write: 'ask',        // create_extraction
  destructive: 'deny',  // delete_extraction
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2

04 · AUTH AND DATA

How Extracta.ai connects.

Users add their Extracta.ai 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 Extracta.ai 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

05 · WORKS WITH

Use Extracta.ai from any agent.

Claude

ChatGPT

Cursor

Codex

VS Code

OpenAI Agents SDK

Claude Agent SDK

Vercel AI SDK

Mastra

LangGraph

06 · RELATED

Extracta.ai and Arc0, answered.

Q01

Can I use Extracta.ai with Claude, ChatGPT or Cursor?

Yes. Connect Extracta.ai 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 Extracta.ai actions you allow.

Q02

How do users connect Extracta.ai?

Users add their Extracta.ai 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.

Q03

Which Extracta.ai actions can my agent take?

10 in total: 4 read, 4 write and 2 destructive, such as “create Extraction”. Your policies decide which of them each agent may call.

Q04

Can I stop my agent from deleting things in Extracta.ai?

Yes. Actions like “delete Extraction” 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.

Q05

Can my own backend call Extracta.ai too?

Yes. The same Extracta.ai connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Extracta.ai into your agent.

Your users connect Extracta.ai once, under your brand. Your agent gets 10 actions behind your policies, with every call on the record.

Connect Extracta.ai Browse all apps →

Free to build · MCP + REST · Audit log on every plan