Anthropic connectors

Integrate Anthropic Claude models into Bonita processes for text generation, classification, data extraction and document analysis with vision.

The Anthropic connector is part of the Bonita AI Connectors family.

Getting started

Import the bonita-connector-ai-anthropic module as an extension dependency in your Bonita project. See the AI connectors overview for general setup instructions.

Connection configuration

Parameter Required Description Default

API Key

Yes

Anthropic API key from Anthropic Console

Resolved from env var AI_API_KEY if not set

Base URL

No

Custom endpoint URL

https://api.anthropic.com/v1

Model Name

No

Claude model to use

claude-sonnet-4-6

Temperature

No

Controls randomness (0.0 to 1.0)

Timeout

No

Request timeout in milliseconds

Available models

  • claude-sonnet-4-6 (default) — Strong at complex analysis, extraction, and generation

  • claude-haiku-4-5-20251001 — Fast and cost-effective for simpler tasks

  • claude-opus-4-6 — Most capable, best for nuanced reasoning

Claude models support native vision for images, making them well suited for document analysis involving visual elements.

See Anthropic Models documentation for the full list.

Operations

Ask

Send a user prompt (with optional system prompt and documents) to a Claude model and return the generated response.

Parameter Required Description Default

User Prompt

Yes

The prompt to send to Claude

System Prompt

No

System instructions to guide the model behavior

You are a polite assistant.

Output JSON Schema

No

JSON Schema to structure the response as JSON

Source Document Reference

No

Bonita process document to include as context

Source Document References

No

List of Bonita process documents to include as context
Parameter Type Description

output

String

The generated response from the model

Classify

Classify a document into one of the predefined categories.

Parameter Required Description Default

Categories

Yes

Comma-separated list of classification categories

Source Document Reference

Yes

Bonita process document to classify

Source Document References

No

List of documents to classify
Parameter Type Description

output

String

JSON with category and confidence fields
Sample classification result
{
  "category": "CONTRACT",
  "confidence": 0.92
}

Extract

Extract structured data from a document using field names or a JSON Schema.

Parameter Required Description Default

Fields to Extract

No

Comma-separated list of field names to extract

Output JSON Schema

No

JSON Schema defining the extraction structure

Source Document Reference

Yes

Bonita process document to extract from

Source Document References

No

List of documents to extract from
You must provide at least one of fieldsToExtract or outputJsonSchema parameters.
Parameter Type Description

output

String

JSON with extracted fields

Use cases

Document analysis with vision

Use Claude’s native vision support to analyze scanned documents, photos of forms, or image-based PDFs with visual elements like signatures, stamps, or handwritten notes.

Process flow:

  1. A scanned identity document is uploaded to the process

  2. A service task uses the Ask connector with the document attached

  3. Claude analyzes both text and visual elements (stamps, signatures, formatting)

  4. The structured result is stored for downstream verification

Configuration:

{
  "apiKey": "${AI_API_KEY}",
  "chatModelName": "claude-sonnet-4-6",
  "systemPrompt": "You are a document analysis expert. Analyze the provided document image and extract all relevant information, including visual elements like signatures and stamps.",
  "userPrompt": "Analyze this scanned identity document. Extract the personal information and indicate whether the document appears authentic (check for stamps, signatures, and standard formatting).",
  "outputJsonSchema": "{\"type\":\"object\",\"required\":[\"documentType\",\"fullName\",\"dateOfBirth\",\"documentNumber\",\"expiryDate\",\"hasSignature\",\"hasOfficialStamp\",\"authenticityAssessment\"],\"properties\":{\"documentType\":{\"type\":\"string\"},\"fullName\":{\"type\":\"string\"},\"dateOfBirth\":{\"type\":\"string\"},\"documentNumber\":{\"type\":\"string\"},\"expiryDate\":{\"type\":\"string\"},\"hasSignature\":{\"type\":\"boolean\"},\"hasOfficialStamp\":{\"type\":\"boolean\"},\"authenticityAssessment\":{\"type\":\"string\"}}}"
}

Expected output:

{
  "documentType": "National ID Card",
  "fullName": "Jean-Pierre Martin",
  "dateOfBirth": "1985-03-15",
  "documentNumber": "FR-1234567890",
  "expiryDate": "2028-03-14",
  "hasSignature": true,
  "hasOfficialStamp": true,
  "authenticityAssessment": "Document appears authentic: official stamp present, standard formatting consistent with French national ID cards, signature field properly completed."
}

Legal contract review

Analyze contracts for GDPR compliance risks, missing clauses, and potential legal issues.

Process flow:

  1. A contract PDF is uploaded or received through a process

  2. A service task uses the Ask connector to analyze the document

  3. The structured findings are stored as BDM objects

  4. A human task presents the analysis for legal review

Configuration:

{
  "apiKey": "${AI_API_KEY}",
  "chatModelName": "claude-sonnet-4-6",
  "systemPrompt": "You are a legal compliance analyst. Identify potential risks, non-compliant clauses, and missing standard provisions.",
  "userPrompt": "Analyze this contract for GDPR compliance issues and list each finding with its severity. Also check for missing standard clauses (liability limitation, termination, force majeure).",
  "outputJsonSchema": "{\"type\":\"object\",\"required\":[\"overallRisk\",\"findings\",\"missingClauses\"],\"properties\":{\"overallRisk\":{\"type\":\"string\"},\"findings\":{\"type\":\"array\",\"items\":{\"type\":\"object\",\"required\":[\"clause\",\"issue\",\"severity\",\"recommendation\"],\"properties\":{\"clause\":{\"type\":\"string\"},\"issue\":{\"type\":\"string\"},\"severity\":{\"type\":\"string\"},\"recommendation\":{\"type\":\"string\"}}}},\"missingClauses\":{\"type\":\"array\",\"items\":{\"type\":\"object\",\"required\":[\"clauseName\",\"importance\",\"recommendation\"],\"properties\":{\"clauseName\":{\"type\":\"string\"},\"importance\":{\"type\":\"string\"},\"recommendation\":{\"type\":\"string\"}}}}}}"
}

Expected output:

{
  "overallRisk": "medium",
  "findings": [
    {
      "clause": "Section 4.2 - Data Processing",
      "issue": "No mention of data subject rights (access, rectification, erasure)",
      "severity": "high",
      "recommendation": "Add explicit provisions for handling data subject requests within 30 days as required by GDPR Article 12."
    },
    {
      "clause": "Section 7.1 - Data Retention",
      "issue": "Retention period of 10 years exceeds necessity for stated purpose",
      "severity": "medium",
      "recommendation": "Reduce retention period or provide justification per GDPR Article 5(1)(e)."
    }
  ],
  "missingClauses": [
    {
      "clauseName": "Data Protection Officer contact",
      "importance": "high",
      "recommendation": "Include DPO contact details as required by GDPR Article 37-39."
    },
    {
      "clauseName": "Force Majeure",
      "importance": "medium",
      "recommendation": "Add standard force majeure clause to protect both parties."
    }
  ]
}

Configuration tips

  • Claude excels at instruction following — use detailed system prompts with specific formatting requirements.

  • The 200K context window makes Claude ideal for analyzing long documents (contracts, reports, transcripts).

  • Native vision support requires no special configuration. Attach image documents and they are analyzed automatically.

  • Use claude-haiku-4-5-20251001 for high-volume classification tasks to reduce costs while maintaining accuracy.

  • Set requestTimeout to at least 60000 ms for complex document analysis tasks.

Source code

bonita-connector-ai on GitHub (module bonita-connector-ai-anthropic)