Integration · Documents

Connect Claude & ChatGPT to Cradl AI

Ask your AI to send a new document to a Cradl AI agent, what an agent run extracted, or which validation tasks are still open. It can create documents, start agent runs and predictions, and read the results. It can also manage agents, models, hooks, actions, validations, users and roles, and read logs and agent statistics.

Connect Cradl AI free

Free plan, no credit card. Takes about a minute.

Try asking

“Send this invoice to our invoice agent.”
The AI creates the document in Cradl AI and starts a run on the agent you name.
“What did the last run of the invoice agent extract?”
It lists the agent's runs and reads the latest one with its results.
“Run the receipts model on document 7b3f.”
It creates a prediction for that document with the model and reads the extracted fields.

What your AI can do

Documents and agents

Create, read, update and delete documents, create and update agents, and start and read agent runs.

Models and predictions

List and manage models, create predictions on documents and read their results.

Validation and admin

Work with validations and their tasks, hooks, actions and functions, and manage users, roles and app clients.

“Which validation tasks are still waiting?”
It lists your validations and the tasks in each one, with their status.
“How many documents did the invoice agent process this month?”
It reads the agent's statistics.
“Why did yesterday's run fail?”
It reads the logs for the workflow execution and shows what happened.

24 tools for Cradl AI

These tools are switched on when you connect. You can switch any of them off, or require your approval before it runs.

archive_agent_runChanges dataArchive an agent run by `agent_id` and `run_id` (status `archived`).
delete_documentDeletesPermanently delete one document by `document_id`. This cannot be undone.
get_agentGet one agent by `agent_id`, including its name, description and linked resources.
get_agent_runGet one agent run by `agent_id` and `run_id`, including its status and extracted results.
get_agent_statisticsGet processing statistics for an agent by `agent_id`.
get_documentGet one document by `document_id`, including its metadata and ground-truth values.
get_modelGet one model by `model_id`, including its field configuration and optional recent statistics.
get_organizationGet your Cradl AI organization's details.
get_predictionGet one prediction by `prediction_id` with the extracted field values and confidence scores.
get_validationGet one validation by `validation_id`.
get_validation_taskGet one review task by `validation_id` and `task_id`, including the input to review.
list_agent_runsList runs for an agent by `agent_id`, optionally filtered by status and created/updated time, with sorting and paging.
list_agentsList your Cradl AI document-processing agents. Use to find an agent ID.
list_documentsList uploaded documents, optionally filtered by dataset ID, with sorting and paging.
list_modelsList the extraction models available to you.
list_predictionsList predictions (extraction results), optionally filtered by model ID, with sorting and paging.
list_usersList users in your Cradl AI organization.
list_validation_tasksList review tasks in a validation by `validation_id`, optionally filtered by status.
list_validationsList validations (human review queues).
run_predictionChanges dataExtract data from an uploaded document by running a model on it: `model_id` and `document_id`, with optional page limit and rotation.
start_agent_runChanges dataStart a new run of an agent by `agent_id`, with optional `variables`. To process a file, upload it afterwards with this run's ID as `agent_run_id`.
update_documentChanges dataUpdate a document's name, description, metadata, dataset, retention or ground-truth values by `document_id`.
update_validation_taskChanges dataComplete or update a review task by `validation_id` and `task_id`: set `status` (e.g. succeeded, failed) and the corrected `output`.
upload_documentChanges dataUpload a document (PDF, JPEG, PNG, TIFF, WebP or email) as base64 `content` with its `content_type`, optionally linked to an agent run via `agent_run_id`, with name and retention days.

Set up in three steps

1
Pick the app
Create a free PipMCP account and choose Cradl AI from the app list.
2
Paste your key
Paste a Cradl AI Client ID and Client secret. In Cradl AI, open your agent and add a new trigger that provides API credentials, for example the n8n trigger. Copy the Client ID and Client Secret shown in the trigger panel.
3
Add the link to your AI
You get a personal MCP link. Add it to Claude, ChatGPT or Cursor:
1. Click your name, then Settings › Connectors › Add custom connector.
2. Paste your link as the Remote MCP server URL.
3. Switch it on from the + menu in a chat.

Questions

What can the AI do in Cradl AI?

It can create, read, update and delete documents, and create, update and run agents. It can create predictions with your models, read results and statistics, and work with validations and validation tasks. It can also manage models, hooks, actions, functions, users, roles and app clients, and read logs.

Does the AI see my Cradl AI credentials?

No. Your Cradl AI Client ID and secret are encrypted at rest and never shown to the AI. After you save them, they are not shown again, not even to you. The AI only sees the results of the tools it calls.

Can I control what the AI is allowed to do?

Yes. You choose which tools are switched on, so you can start with documents, runs and results only. Deleting documents and changing roles are permanent or affect who has access, so they can require your approval before they run. Every tool call is logged.

Does it work with ChatGPT?

Yes. In ChatGPT go to Settings › Apps & Connectors › Advanced and turn on Developer mode, then add your PipMCP link. Developer mode needs a paid ChatGPT plan: Plus, Pro, Business or Enterprise. The same link also works in Claude (Settings › Connectors › Add custom connector), Cursor and other MCP clients.

What does it cost?

PipMCP has a free plan with no credit card required. Paid plans bill per completed task. You also need your own Cradl AI account.

Where do I find my Cradl AI Client ID and secret?

Cradl AI shows API credentials on an agent's triggers. Open the agent, add a trigger that provides API credentials, such as the n8n trigger, and copy the Client ID and Client Secret from the trigger panel. PipMCP uses them to get an access token from Cradl AI for each session.