test-reasoning

Verify ReasoningConfig serialization into provider-specific JSON schemas.

7.5k|1.4k|Updated Dec 8, 2024
One-click install
npx skills add https://github.com/tailcallhq/forgecode --skill test-reasoning
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: test-reasoning
Source: https://github.com/tailcallhq/forgecode/tree/main/.forge/skills/test-reasoning
Command: npx skills add https://github.com/tailcallhq/forgecode --skill test-reasoning

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Validate that ReasoningConfig fields are correctly serialized into provider-specific JSON for OpenRouter, Anthropic, GitHub Copilot, and Codex.

Core Features & Use Cases

  • Automated tests verify correct mapping of reasoning parameters across multiple AI providers.
  • Validates edge cases such as enabled, exclude, and max_tokens across provider schemas.
  • Use case: ensure that production-grade reasoning configurations are serialized as expected before deployment.

Quick Start

Run the bundled script to execute all reasoning-serialization tests across providers.

Frequently Asked Questions about test-reasoning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate reasoning serialization across multiple AI providers?

You can validate reasoning serialization across AI providers by running automated tests that verify correct JSON mapping of ReasoningConfig fields for OpenRouter, Anthropic Claude, GitHub Copilot, and Codex models.

What reasoning config fields are covered during provider serialization testing?

Provider serialization testing covers ReasoningConfig fields including effort, max_tokens, enabled, and thinking budget, verifying their correct mapping across various scenarios like enabled and exclude tests.

How do I test edge cases for AI reasoning parameters before deployment?

You can test edge cases for AI reasoning parameters by executing bundled scripts that validate enabled, exclude, and combination tests, ensuring production-grade reasoning configurations serialize correctly before deployment.

Does this serialization testing support OpenRouter, Anthropic, GitHub Copilot, and Codex?

Yes, this serialization testing supports OpenRouter, Anthropic Claude, GitHub Copilot, and Codex models, verifying that reasoning configurations map correctly to each specific provider's JSON schema.

Why do my reasoning configuration tests fail when mapping to provider schemas?

Reasoning configuration tests fail to surface misconfigurations in output JSON mapping to provider schemas, ensuring that incorrect serialization of fields like max_tokens or thinking budget is caught before deployment.