Agent Skills: Test Runner Skill

Test Runner Skill

Overview

The Skill(foundry:run-tests) skill provides systematic test debugging with a 5-phase investigation workflow. It guides debugging complex test failures and leverages AI consultation when available.

When to use this skill:

• Debugging test failures that aren't obvious
• Systematic investigation needed (flaky tests, complex failures)
• AI consultation would help diagnose the issue

When NOT to use this skill:

• Simple test runs (just use pytest, go test, npm test directly)
• Tests are passing
• Failure cause is obvious

Core Workflow

[x?]=decision · (GATE)=user approval · →=sequence · ↻=loop

- **Entry** → Run failing test(s)
  - [pass?] → **Exit**: Done
  - [fail?] → Categorize failure
- FormHypothesis → GatherContext[Explore preferred]
- [Research available?] → `research action="chat"`
  - [yes] → Consult (MANDATORY)
  - [no] → skip
- ImplementFix → VerifySpecific ↻ [pass?]
- RunFullSuite ↻ [pass?] → **Exit**: Done

Decision rules:

• Tests pass? Done.
• Simple fix (typo/obvious)? Fix, then verify.
• Complex/unclear? Investigate, consult AI, fix, verify.

Phase 1: Run Tests

Use native test commands:

# Python
pytest tests/test_module.py::test_function -vvs

# Go
go test -v -run TestName ./...

# JavaScript
jest path/to/test.js -t "test name"
npm test

Phase 2: Investigate Failures

Categorize the failure:

| Category | Description | |----------|-------------| | Assertion | Expected vs actual mismatch | | Exception | Runtime errors (language-specific) | | Import | Missing dependencies or module issues | | Setup | Fixture, configuration, or initialization issues | | Timeout | Performance or hanging issues | | Flaky | Non-deterministic failures |

Extract key information:

• Test file and function name
• Line number where failure occurred
• Error type and message
• Full stack trace

Form hypothesis: What's causing the failure?

For detailed failure categories, see references/failure-categories.md

Phase 3: Gather Code Context

Use Explore subagents (preferred) for code context, or Glob, Grep, and Read for targeted lookups.

Subagent selection:

• Explore (quick) - Find related test files
• Explore (medium) - Understand module dependencies
• Explore (very thorough) - Multi-file state/fixture investigation

For detailed subagent patterns, see references/subagent-patterns.md

Phase 4: AI Consultation

Mandatory for complex failures when research tools are available.

Standard consultation (most cases):

mcp__plugin_foundry_foundry-mcp__research action="chat" prompt="..." system_prompt="You are debugging a test failure."

Multi-perspective analysis (flaky tests, architectural issues):

mcp__plugin_foundry_foundry-mcp__research action="consensus" prompt="..." strategy="synthesize"

When to use which:

• chat - Standard debugging, quick diagnosis
• consensus - Complex failures, want multiple AI perspectives, flaky tests
• thinkdeep - Systematic hypothesis-driven investigation for complex issues

CRITICAL: Read references/tool-selection.md before AI consultation. Contains required prompt templates and strategy parameters.

Phase 5: Fix & Verify

1. Synthesize findings from investigation + AI recommendations
2. Implement fix using Edit tool
3. Verify with specific test (native command)
4. Run full suite to check for regressions

Requirement Discovery (SDD Integration)

When debugging reveals missing spec requirements, document them:

mcp__plugin_foundry_foundry-mcp__task action="add-requirement" spec_id={spec-id} task_id={task-id} requirement="Handle empty input array gracefully"

When to use:

• Test failure reveals undocumented edge case
• Fix requires behavior not in acceptance criteria
• Investigation uncovers missing validation rule

Runner-Specific Notes

Python (pytest)

• Run specific: pytest tests/test_file.py::test_name -vvs
• Debug mode: pytest --pdb tests/test_file.py::test_name
• Markers: -m "unit", -m "not slow"
• Coverage: pytest --cov=src --cov-report=term-missing

Go

• Run specific: go test -v -run TestName ./...
• Race detection: go test -race ./...
• Coverage: go test -cover -coverprofile=coverage.out ./...
• Debug: dlv test -- -test.run TestName

JavaScript (Jest)

• Run specific: jest path/to/test.js -t "test name"
• Watch mode: jest --watch
• Debug: node --inspect-brk node_modules/.bin/jest --runInBand
• Coverage: jest --coverage

npm test

• Standard: npm test
• Pass args: npm test -- --specific-flag

Detailed Reference

• Failure categories → references/failure-categories.md
• Investigation patterns → references/investigation.md
• Subagent patterns → references/subagent-patterns.md
• AI prompt templates → references/tool-selection.md
• Common fixes → references/common-fixes.md
• Troubleshooting → references/troubleshooting.md