AI Agent Configuration

The following section outlines the steps required to configure the AI agent.

Agent Settings

• Navigate to Global Settings -> Search for AI Agent Settings and click on Configure Settings action item.
• Update the following settings based on the requirement -

Settings key	Description
AI_AGENT_API_KEY	The authentication key issued by the AI agent provider.
AI_PROVIDER	The name of the AI provider (e.g., ANTHROPIC, OPENAI).
AI_MODEL	The specific model of the configured AI PROVIDER that will be used for code generation.
VECTOR_DB_URL	The endpoint URL of the vector database service.
VECTOR_DB_TOKEN	The authentication token used to access the vector database.
VECTOR_DB_SEARCH_TOKEN	The token used specifically for search/query operations in the vector database.
MULE_VECTOR_DB_COLLECTION_NAME	The name of the vector database collection containing MuleSoft related embeddings.
API_VECTOR_DB_COLLECTION_NAME	The name of the vector database collection containing API related embeddings..
MULE_AGENT_VALIDATOR_CMD	The maven command to be executed for validating the project.
MAVEN_SETTINGS_XML	The Maven settings.xml configuration used for dependency resolution and repository authentication.
MULE_SPECIFICATIONS	Organization specific guidelines and standards that the AI agent must follow while generating the mule project.
API_SPECIFICATIONS	Organization specific guidelines and standards that the AI agent must follow while generating the API specification.
DOC_SPECIFICATIONS	Organization specific guidelines and standards that the AI agent must follow while generating the documentation.
<LANGUAGE>_TEMPLATE_URL	Location of the initial template to be used for generating the project. For Example - MULE_API_PROJECT_TEMPLATE_URL represents the template to be used for generating the mule project. This setting is application only while generating new projects and not while editing an existing project.

• Save the configuration.

See Also

• Automation Tasks
• Automation Task Logs