Skip to the operations

AI

Anthropic in a decision flow.

4 Anthropic operations a decision flow can call directly, with the credentials your own contract issued. The response is data the rest of the flow reads, branches on, and keeps in the trace.

Category
AI
Type
Integration
Authentication
API key
Test environment
Production host only

Who they are

The AI lab behind the Claude models.

Anthropic is an AI safety company that builds the Claude family of large language models and sells access to them through its own API. The models are used for reading and summarising documents, pulling structure out of free text, and drafting language a person will read — the work that sits outside a scorecard. Claude is available directly and through the major cloud providers.

What a flow can call

4 operations, each one a step you can place on the canvas.

  1. POST/v1/messages

    Ask a model (Messages)

    One request, one answer. max_tokens is required and is a hard ceiling on the reply — if it is hit the response comes back truncated with stop_reason "max_tokens" rather than failing.

  2. POST/v1/messages

    Extract structured JSON

    Same endpoint, constrained to a JSON schema so the reply parses without a regex.

  3. POST/v1/messages/count_tokens

    Count tokens before sending

    Returns the input token count for a request without running it.

  4. GET/v1/models

    List available models

    Model ids, context windows and capabilities as the account sees them.

Where it sits in the decision

AI calls have a natural place in a flow.

A model call is for the judgment a table cannot express — reading a document, summarising a file, weighing a narrative. Because the output is not deterministic, a flow constrains it with a schema and keeps the prompt and the response in the trace, so the reasoning behind a referral stays as inspectable as a rule.

Whatever Anthropic returns is part of the run, so it is part of the record. When someone asks months later why an applicant was declined, the answer cites what came back at the time rather than re-fetching from a service whose answer has since changed.

  • 01Add Anthropic as a connection authenticating with an API key.
  • 02Anthropic has one host for both environments, so guard test runs with your own credentials and limits.
  • 03Place a Connection node and pick an operation — “Ask a model (Messages)” is usually the first one a flow needs.
  • 04Map the response into the fields your rules read, then test the whole path before it carries live traffic.

Common questions

Using Anthropic in a flow.

How do I connect Anthropic to a decision flow?

Add Anthropic as a connection in your workspace authenticating with an API key, with the credentials your own contract issued — ArboRule calls the provider as you, and never holds a contract on your behalf. Once the connection exists, any flow in the workspace can place a Connection node and choose one of its operations. The credentials live on the connection, not in the flow, so a policy owner can use Anthropic in a decision without ever seeing the secret.

Can I test Anthropic without touching production?

Anthropic exposes one host for both environments, so there is no separate sandbox to point at. Test runs still execute in Sandbox and are recorded separately in decision history, but the call goes to the same place as production — so guard it with your own test credentials, rate limits, or data.

What can a flow call on Anthropic?

4 operations, including “Ask a model (Messages)”, “Extract structured JSON”, “Count tokens before sending”. Each one is a step you place on the canvas and map into the fields your rules read, and most flows start with “Ask a model (Messages)”. The list comes from the same manifest the engine uses to make the call, so this page cannot describe an operation the product does not have.

Where in a decision should Anthropic be called?

A model call is for the judgment a table cannot express — reading a document, summarising a file, weighing a narrative. Because the output is not deterministic, a flow constrains it with a schema and keeps the prompt and the response in the trace, so the reasoning behind a referral stays as inspectable as a rule.

Also in AI

Ready when you are

Wire Anthropic into a real decision.

Build the flow in Sandbox, connect your account, and watch the decision pull what it needs before it answers.

Read the docs