test-agent

name	test-agent
description	Write comprehensive test suites covering happy paths, error paths, edge cases, and integration points. Use when adding test coverage, writing regression tests for bugs, or building e2e tests for critical flows.
domain	agents
tags	["agent","ai-agent","automation","orchestration","test"]

Test Agent

When to Use

Trigger phrases:

• "test agent"

• "Writing tests for new features before or after implementation"

• "Adding missing test coverage for existing code"

• "Writing regression tests for reported bugs"

• Writing tests for new features before or after implementation

• Adding missing test coverage for existing code

• Writing regression tests for reported bugs

• Creating integration tests for API endpoints

• Building end-to-end tests for critical user flows

• Improving test coverage metrics (meaningfully, not just line counting)

• Setting up test infrastructure for a new project

When NOT to Use

• When the task is simple enough for a single command
• When real-time human judgment is required
• When the agent lacks access to required tools or data

Overview

Test Agent is an AI agent skill for agent orchestration. It enables autonomous execution of complex tasks with minimal human intervention.

Capabilities

• Autonomous operation — Execute multi-step test agent workflows independently
• Context awareness — Adapt behavior based on current state and history
• Error recovery — Handle failures gracefully with retry and fallback logic
• Integration — Connect with external tools and services as needed

Workflow

# Example: Agent orchestration
from dataclasses import dataclass

@dataclass
class Task:
    name: str
    priority: int
    assigned_agent: str

def orchestrate(tasks: list[Task]) -> dict:
    results = {}
    for task in sorted(tasks, key=lambda t: t.priority):
        results[task.name] = execute(task)
    return results

1. Initialize — Set up the agent context and load required resources
2. Plan — Break down the task into executable steps
3. Execute — Run each step, monitoring for errors and adapting as needed
4. Verify — Validate results against acceptance criteria
5. Report — Summarize outcomes and suggest next steps

Configuration

• Define task objectives and constraints clearly
• Set appropriate timeout and retry limits
• Configure tool access and permissions
• Enable logging for debugging and audit

Anti-Rationalization

Rationalization	Reality
"I will just do it manually"	Agents automate repetitive tasks — manual work does not scale
"The agent will figure it out"	Without clear instructions, agents hallucinate. Give explicit context.
"One agent is enough"	Complex tasks benefit from specialized agents working in parallel

Process

1. Design — Define interface, identify patterns, plan implementation
2. Implement — Write code following existing conventions, add tests
3. Verify — Run tests, check integration, validate behavior

Verification

• All steps executed successfully
• Results validated against acceptance criteria
• Error handling tested with edge cases
• Documentation updated with findings