agentic-testing

Define behavior-focused and contract testing best practices for agentic LLM coding.

1|Updated Mar 16, 2026
One-click install
npx skills add https://github.com/otography/otography-app --skill agentic-testing-otography
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agentic-testing
Source: https://github.com/otography/otography-app/tree/main/.claude/skills/agentic-testing
Command: npx skills add https://github.com/otography/otography-app --skill agentic-testing-otography

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Agentic coding often lacks structured testing guidance, risking unstable iterations and brittle agents. This skill provides a principled testing framework to guide tests, reviews, and CI workflows.

Core Features & Use Cases

  • Behavior-focused testing guidance for agentic code
  • Layered testing approach (Deterministic Foundations, Contract Tests, Behavior Tests)
  • Best-practice principles (R1-R6) for fast, reliable iteration

Quick Start

Follow a guided agentic testing workflow to validate behavior-focused tests.

Frequently Asked Questions about agentic-testing

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

FAQPage Schema
How do I write behavior-focused tests for agentic LLM coding workflows?

Behavior-focused tests for agentic LLM coding validate high-level outcomes by applying best-practice principles that ensure deterministic, fast tests with actionable failures, avoiding brittle implementation details.

What is the best way to structure contract testing for iterative LLM agent development?

Contract testing for iterative LLM agent development establishes strict boundaries between internal modules and external mocks, ensuring reliable component integration and supporting stable iterative refactoring without breaking existing workflows.

How does test-driven development work when refactoring agentic code?

Test-driven development for refactoring agentic code applies a layered testing approach—deterministic foundations, contract tests, and behavior tests—to validate logic before changes, ensuring reliable iteration and preventing unstable agent behavior.

Can I use this testing framework to plan test coverage and CI testing for LLM agents?

Yes, you can apply this framework to plan test coverage and CI testing for LLM agents by following guided workflows that enforce fast, reliable test execution and actionable failure reporting during continuous integration.

Why do my LLM agent tests keep failing during iterative development?

LLM agent tests fail during iterative development when they test implementation details rather than behavior, or lack deterministic foundations; applying principled testing boundaries and external mocks resolves these brittle failures.

Do I need to mock external modules to run deterministic tests for agentic coding?

Yes, mocking external modules is required to run deterministic tests for agentic coding, establishing strict boundaries that isolate internal agent logic from external dependencies and ensure fast, reliable test execution.