home / skills / laurigates / claude-plugins / test-consult

This skill consults the test-architecture agent to shape testing strategy and design for new features, coverage gaps, and flaky tests.

npx playbooks add skill laurigates/claude-plugins --skill test-consult

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
2.2 KB
---
model: opus
created: 2025-12-16
modified: 2025-12-16
reviewed: 2025-12-16
allowed-tools: Task, Read, Glob, Grep
argument-hint: "<topic> [--context <path>]"
description: Consult test-architecture agent for testing strategy and design
name: test-consult
---

## Context

Explicit command to consult the `test-architecture` agent for strategic testing decisions.

**Use when:**
- Creating tests for NEW features
- Test strategy needs adjustment
- Coverage gaps identified
- Flaky tests detected
- Framework selection questions

## Topics

| Topic | What it covers |
|-------|----------------|
| `coverage` | Analyze gaps, recommend high-value tests to write |
| `strategy` | Review overall test architecture and pyramid |
| `framework` | Recommend testing tools for tech stack |
| `flaky` | Diagnose and remediate flaky tests |
| `pyramid` | Analyze test tier distribution |
| `new-feature` | Design test approach for new functionality |
| (free text) | Custom question about testing |

## Parameters

- `$1`: Topic keyword or natural language question
- `--context <path>`: Focus analysis on specific directory

## Behavior

1. **Gather Project Context**:
   - Find test directories and structure
   - Detect frameworks in use (pytest, vitest, playwright, etc.)
   - Read coverage reports if available
   - Check recent test failures

2. **Delegate to test-architecture Agent**:
   - Provide gathered context
   - Include specified topic
   - Request actionable recommendations

3. **Present Recommendations**:
   - Strategic advice
   - Specific action items
   - Offer to create issues/tasks

## Examples

```bash
# Coverage analysis
/test:consult coverage

# New feature testing strategy
/test:consult new-feature --context src/auth/

# Framework decision
/test:consult "Should we add property-based testing?"

# Flaky test investigation
/test:consult flaky

# Custom question
/test:consult "How should we structure tests for our API endpoints?"
```

## Agent Delegation

This command delegates to the `test-architecture` agent with:
- Project test structure analysis
- Coverage data (if available)
- Specified topic/question
- Context path (if provided)

The agent will provide:
- Strategic recommendations
- Specific actionable items
- Trade-off analysis
- Implementation guidance

Overview

This skill consults the test-architecture agent to produce strategic, actionable guidance for testing strategy and design. It helps teams choose frameworks, close coverage gaps, diagnose flaky tests, and design testing approaches for new features. Use it to turn test findings into prioritized, implementable tasks.

How this skill works

The skill scans the repository for test folders, detects testing frameworks, and reads available coverage and recent failure data. It forwards that context plus the specified topic or question to the test-architecture agent. The agent returns strategic recommendations, concrete action items, remediation steps, and optional issue/task suggestions.

When to use it

  • Designing tests for a new feature or API
  • Reviewing and adjusting overall test strategy or pyramid
  • Identifying and prioritizing coverage gaps to write high-value tests
  • Investigating and fixing flaky or intermittent tests
  • Choosing or evaluating test frameworks and tooling for your stack

Best practices

  • Provide a context path when you want focused analysis on a specific module or service
  • Include recent failing test logs or coverage reports to get precise recommendations
  • Ask for prioritized action items to convert strategy into tickets or pull requests
  • Use topic keywords (coverage, strategy, flaky, framework, new-feature) for targeted advice
  • Iterate: run recommended changes, attach results, and re-consult for follow-up

Example use cases

  • Analyze test coverage for the payments service and get a prioritized list of tests to add
  • Get a testing approach and test types for a newly planned auth flow under src/auth/
  • Diagnose intermittent failures after CI migration and receive remediation steps
  • Evaluate whether adding property-based or contract testing fits the current tech stack
  • Request a recommended distribution across unit, integration, and E2E tests (test pyramid)

FAQ

What input does the skill require?

Pass a topic keyword or a natural-language question as $1; use --context <path> to focus on a directory.

Can it create issues or tasks?

Yes. The agent can produce actionable items and can format them as issues or tasks for your tracker on request.