Test Framework Detection

What this skill does

# Test Framework Detection

**CRITICAL: The description field above controls when Claude auto-loads this skill.**

## Overview

Provides detection scripts and patterns for identifying installed test frameworks, coverage tools, CI workflows, and testing infrastructure across JavaScript/TypeScript, Python, Go, and Rust projects.

## Instructions

### 1. Detection Strategy

Detection uses a **confidence scoring** approach (mirrors `stack-detector` pattern):
- **Evidence file exists** + **dependency declared** = high confidence match
- **Evidence file only** = medium confidence (may be leftover)
- **Dependency only** = low confidence (installed but not configured)

### 2. Running Detection

**Full detection orchestrator:**
```bash
bash scripts/detect-test-frameworks.sh [project-path]
```

**Language-specific detection:**
```bash
bash scripts/detect-node-testing.sh [project-path]    # Jest, Vitest, Mocha, AVA
bash scripts/detect-python-testing.sh [project-path]   # pytest, unittest, nose2
bash scripts/detect-go-testing.sh [project-path]       # go test, testify
bash scripts/detect-rust-testing.sh [project-path]     # cargo test, criterion
bash scripts/detect-ci-workflows.sh [project-path]     # GitHub Actions, GitLab CI
```

### 3. Detection Patterns

**JavaScript/TypeScript:**
| Framework | Evidence Files | Dependencies |
|-----------|---------------|--------------|
| Vitest | `vitest.config.ts`, `vitest.config.js` | `vitest` in devDependencies |
| Jest | `jest.config.js`, `jest.config.ts`, `jest.config.mjs` | `jest` in devDependencies |
| Mocha | `.mocharc.yml`, `.mocharc.json`, `.mocharc.js` | `mocha` in devDependencies |
| AVA | `ava.config.js`, `ava.config.cjs` | `ava` in devDependencies |
| Playwright | `playwright.config.ts`, `playwright.config.js` | `@playwright/test` in devDependencies |

**Python:**
| Framework | Evidence Files | Dependencies |
|-----------|---------------|--------------|
| pytest | `pytest.ini`, `pyproject.toml` [tool.pytest], `setup.cfg` [tool:pytest], `conftest.py` | `pytest` in requirements/pyproject |
| unittest | `test_*.py` files with `import unittest` | stdlib (always available) |
| nose2 | `unittest.cfg`, `setup.cfg` [unittest] | `nose2` in requirements |

**Go:**
| Framework | Evidence Files | Dependencies |
|-----------|---------------|--------------|
| go test | `*_test.go` files | stdlib (always available) |
| testify | `*_test.go` with `testify` imports | `github.com/stretchr/testify` in go.mod |

**Rust:**
| Framework | Evidence Files | Dependencies |
|-----------|---------------|--------------|
| cargo test | `#[cfg(test)]` modules, `tests/` directory | stdlib (always available) |
| criterion | `benches/` directory | `criterion` in Cargo.toml dev-dependencies |

### 4. Coverage Tool Detection

| Tool | Language | Evidence |
|------|----------|----------|
| v8/c8 | Node.js | `c8` in devDependencies or `--coverage` in test script |
| istanbul/nyc | Node.js | `nyc` in devDependencies |
| coverage.py | Python | `coverage` in requirements, `.coveragerc` |
| go cover | Go | Built-in, `go test -cover` |
| tarpaulin | Rust | `cargo-tarpaulin` installed |

### 5. Output Format

Detection writes an expanded `testing` key to `.claude/project.json`:

```json
{
  "testing": {
    "unit": {
      "framework": "vitest",
      "config": "vitest.config.ts",
      "command": "npm run test"
    },
    "api_contract": {
      "framework": "newman",
      "collections": [],
      "auth_strategy": "none"
    },
    "e2e": {
      "framework": "playwright",
      "config": "playwright.config.ts"
    },
    "smoke": {
      "health_endpoint": "/api/health",
      "critical_paths": ["/", "/login"]
    },
    "coverage": {
      "tool": "v8",
      "threshold": 80
    },
    "ci_workflows": [".github/workflows/test.yml"],
    "ai_detected": false,
    "llm_evals_plugin": "llm-evals"
  }
}
```

### 6. CI Workflow Detection

Scans `.github/workflows/*.yml` and `.gitlab-ci.yml` to identify:
- Which test types are already running in CI
- Test commands used in CI vs local
- Coverage reporting in CI
- Test result artifact storage

### 7. AI Framework Detection

Flags `ai_detected: true` when any of these are found:
- `@ai-sdk/` packages in dependencies
- `langchain` in dependencies
- `openai` SDK in dependencies
- `@anthropic-ai/sdk` in dependencies
- `transformers` in Python dependencies

When AI is detected, notes that LLM eval testing should use the `llm-evals` plugin.

## Available Scripts

1. **detect-test-frameworks.sh** - Main orchestrator that runs all detectors
2. **detect-node-testing.sh** - Detect Jest, Vitest, Mocha, AVA, Playwright for Node.js
3. **detect-python-testing.sh** - Detect pytest, unittest, nose2 for Python
4. **detect-go-testing.sh** - Detect go test, testify for Go
5. **detect-rust-testing.sh** - Detect cargo test, criterion for Rust
6. **detect-ci-workflows.sh** - Detect GitHub Actions and GitLab CI test workflows

## Available Templates

1. **project-json-testing-schema.json** - JSON schema for the expanded testing key
2. **detection-report.md** - Template for detection result reporting

## Available Examples

1. **node-project-detection.md** - Example detection for a Node.js/TypeScript project
2. **python-project-detection.md** - Example detection for a Python project
3. **multi-stack-detection.md** - Example detection for a multi-language project

## Requirements

- bash shell for running detection scripts
- jq for JSON processing
- Access to project root directory

## Progressive Disclosure

- Read `examples/node-project-detection.md` for Node.js detection walkthrough
- Read `examples/python-project-detection.md` for Python detection walkthrough
- Read `templates/project-json-testing-schema.json` for the complete schema definition