Validator Console Guide

This page provides a guide to the Validator entity in the Namirasoft Inference Console. It defines the concepts and configuration fields used when creating and managing Validators. Use this guide to understand the purpose and behavior of each setting available during Validator configuration.

What Is a Validator?

A Validator confirms that an AI response arrives in the structure you expect: JSON, YAML, ENV, XML, a number, a date, a time, a date and time, or one value from a list you define. When a response is invalid, it either returns a clear error or corrects the output automatically by calling another inferencer, with a retry limit you control.

The Challenge a Validator Solves

Models answer in prose unless something enforces structure. Applications break on responses they cannot parse.

Common challenges include:

• Unparseable output: A response with explanatory text around the JSON fails parsing. The failure surfaces in production, not in review.
• Retry code everywhere: Each application writes its own detect-and-retry loop for malformed output. The loops differ, and none of them is anyone’s job.
• Silent drift: A model change can quietly alter the output shape. Nothing catches it until something downstream breaks.

How Namirasoft Inference Solves the Problem

You choose the format Type, define the expected structure, and decide what happens on invalid output. With Error, the request fails with the message you define. With Correct, the corrector inferencer you choose rewrites the response into the expected format, up to your retry count. Attach the Validator to any inferencer, or to a Gateway, and every response it returns is checked.

For example, a Validator named “JSON Response Validator” with Type JSON keeps every response parseable for your application, and the Correct action can fix a response that arrives with extra text around the JSON.

Overview of Validator Fields and Options

Below is a detailed explanation of the fields available when creating or managing a Validator. Understanding these fields helps ensure your Validator is configured correctly for your requirements.

• ID (String): This is a unique identifier automatically assigned to the Validator when it is created. The system uses it to track, reference, and manage this specific Validator. This value is auto-generated and cannot be modified.

• User ID (Namirasoft Account’s ID): This is the unique identifier of the Namirasoft Account user who owns this Validator. It is used internally for permission control, audit logging, and access management.

• Workspace ID (Namirasoft Workspace’s ID): This is the identifier of the workspace this Validator belongs to, as defined in Namirasoft Workspace. A workspace is a shared organizational space where teams group their work, projects, and members.

• Name (String): This is a label used to identify this Validator in the console. A good name states the format it enforces. For example: “Order JSON Validator”.

• Type (Enum): This is the format the response must arrive in. The options are JSON, YAML, ENV, XML, Enum, Number, Date, DateTime, and Time.

• Expected Format (String): This is the structure the response is checked against, written in the selected type. For example, for JSON you can define the fields and value types the response must contain.

• Example Format (String): This is an example of a valid response, and it is optional. A concrete example helps both the validation and the automatic correction produce exactly the shape you want.

• On Invalid Action (Enum): This decides what happens when a response does not match the expected format. Error rejects the response and returns your error message. Correct sends the response to the corrector inferencer to be fixed automatically.

• Error Message (String): This is the message returned when the action is Error and validation fails. Write it for the developer who will read it, for example: “Response must be valid order JSON.”

• Corrector Inferencer (Inferencer Type and ID): This is the inferencer that fixes invalid output when the action is Correct. Any inferencer can be the corrector, and it receives the invalid response together with your expected format.

• Retry Count (Integer): This is the maximum number of correction attempts when the action is Correct. When the limit is reached without a valid response, the request fails, so correction never runs away.

• Additional Prompt (String): This is optional extra guidance added to the validation and correction requests. Use it to state rules the format alone cannot express.

• Memory (Memory’s ID): This determines whether the Validator keeps conversation context across requests. When a Memory is connected, relevant conversation history is stored and provided as context when the Validator handles later requests.

• Wiper (Wiper’s ID): This determines whether stored data associated with the Validator is cleaned up according to a defined policy. A Wiper can remove data based on time or count limits.

• Log Group (Log Group’s ID): This determines where the Validator’s execution logs are recorded. When a Log Group, defined in Namirasoft Log, is connected, request activity and errors are recorded there for debugging and review.

• Description (String): This is optional explanatory text about this Validator. Use it to describe the format it enforces and where it is used.

• Created At (DateTime): This is the date and time when this Validator was created. This value is automatically generated and cannot be modified.

• Updated At (DateTime): This is the date and time when this Validator configuration was last modified. This value is updated automatically whenever any field is changed.