AI Agents / Tool contracts and execution
AI Agents: Required argument schemas
Declare required fields, types and allowed values and validate them before dispatch.
Environment and inputs: Model tool calling with a validated application dispatcher; Ollama is a local-runtime example.
When to use it
The model proposes a tool call; application code decides whether and how it executes. Test the contract at the dispatcher boundary so a fluent model response cannot conceal an invalid argument or an uncommitted operation.
Practical example
Illustrative JSON Schema for a lookup-only tool. Ownership must still be checked by the authenticated executor.
{
"type": "object",
"properties": {
"ticket_id": {
"type": "string",
"minLength": 1
}
},
"required": [
"ticket_id"
],
"additionalProperties": false
}Check the result
Submit missing fields, unexpected properties and wrong types and verify rejection before execution.
Save and compare results
Save the test input, configuration and result. Fix the failing boundary, repeat the same test and compare it with the saved working configuration.
Scope and limits
These exercises have not been tested on your systems. Check feature support and configuration names against the installed release and linked documentation.