Cradl AI for AI agents

Cradl AI extracts structured data from documents using AI, letting a team turn invoices or forms into usable data without manual entry. 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 Cradl AI See the 11 actions ↓

DocumentsOAuth 2.0 client credentialsMCP + RESTcradl.ai

AUDIT LOG · CRADL AIPOLICY: acme-support

09:41:07 · claude · u_8f2read

cradl_ai.list_agents

List Agents✓ allowed · 212ms

09:41:08 · claude · u_8f2read

cradl_ai.get_document

Get Document✓ allowed · 164ms

09:41:09 · claude · u_8f2write

cradl_ai.start_agent_run

Start Agent Run✓ approved · approved by user

EVERY CRADL AI CALL, ON THE RECORD

01 · USE CASES

What agents do in Cradl AI.

Process a document for extraction

Process a document through a workflow to pull out its structured fields.

Check a prediction's results

Check a prediction's extracted fields before they get pushed into another system.

Start an agent run on a batch

Start an agent run across a batch of documents instead of processing them one by one.

02 · ACTIONS

11 Cradl AI actions, graded by risk.

Every Cradl 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.

cradl_ai.list_agents
List Agents
cradl_ai.get_document
Get Document
cradl_ai.get_agent_run
Get Agent Run
cradl_ai.get_prediction
Get Prediction
cradl_ai.list_documents
List Documents
cradl_ai.list_agent_runs
List Agent Runs
cradl_ai.list_predictions
List Predictions
cradl_ai.download_file
Download File

write

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

cradl_ai.start_agent_run
Start Agent Run
cradl_ai.process_document
Process Document
cradl_ai.upload_document_to_run
Upload Document to Run

destructive

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

No destructive actions.

03 · HOW IT WORKS

Cradl AI in three steps.

01Your users connect Cradl AIThey add their Cradl AI oauth 2.0 client credentials on Arc0 Connect, under your brand. It goes straight into the vault.
02You set the rulesReads run, and writes like “start Agent Run” can wait for the user to approve.
03Any agent can actYour agent calls Cradl 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('cradl_ai', {
  read: 'allow',
  write: 'ask',        // start_agent_run
  destructive: 'deny',  
})

# 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 Cradl AI connects.

Connect Cradl AI with a service credential for your workspace. Arc0 keeps it in the vault, exchanges it for short-lived tokens, and never passes it to the model.

The same Cradl 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: OAuth 2.0 client credentials
CREDENTIALS: Per-tenant encrypted vault
MODEL SEES: Results only, never credentials
AUDIT LOG: Every call, on every plan

05 · WORKS WITH

Use Cradl AI from any agent.

Claude

ChatGPT

Cursor

Codex

VS Code

OpenAI Agents SDK

Claude Agent SDK

Vercel AI SDK

Mastra

LangGraph

06 · RELATED

Cradl AI and Arc0, answered.

Q01

Can I use Cradl AI with Claude, ChatGPT or Cursor?

Yes. Connect Cradl 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 Cradl AI actions you allow.

Q02

How do users connect Cradl AI?

Connect Cradl AI with a service credential for your workspace. Arc0 keeps it in the vault, exchanges it for short-lived tokens, and never passes it to the model.

Q03

Which Cradl AI actions can my agent take?

11 in total: 8 read, 3 write and 0 destructive, such as “start Agent Run”. Your policies decide which of them each agent may call.

Q04

Can I make my agent read-only in Cradl AI?

Yes. Allow read actions and deny writes in the Cradl AI policy. Your agent can still look things up, and any write it attempts is blocked and logged.

Q05

Can my own backend call Cradl AI too?

Yes. The same Cradl 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 Cradl AI into your agent.

Your users connect Cradl AI once, under your brand. Your agent gets 11 actions behind your policies, with every call on the record.

Connect Cradl AI Browse all apps →

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