managing-pytest

name	managing-pytest
description	Use when working with Pytest — pytest testing framework management and analysis. Covers test discovery, fixture analysis, coverage reporting, marker-based filtering, parametrized test review, plugin management, and conftest configuration. Use when managing pytest suites, investigating test failures, reviewing code coverage, or analyzing test performance.
connection_type	pytest
preload	false

Pytest Testing Framework Management Skill

Manage and analyze pytest test suites, coverage, and configurations.

Core Helper Functions

#!/bin/bash

# Run pytest with JSON output
pytest_run() {
    local args="${1:-}"
    python -m pytest $args --tb=short -q 2>&1
}

# Parse pytest JUnit XML results
pytest_results() {
    local report="${1:-report.xml}"
    python -c "
import xml.etree.ElementTree as ET
tree = ET.parse('$report')
root = tree.getroot()
import json
print(json.dumps({
    'tests': int(root.attrib.get('tests', 0)),
    'failures': int(root.attrib.get('failures', 0)),
    'errors': int(root.attrib.get('errors', 0)),
    'skipped': int(root.attrib.get('skipped', 0)),
    'time': float(root.attrib.get('time', 0))
}))
" 2>/dev/null
}

MANDATORY: Discovery-First Pattern

Always discover pytest configuration and test structure before querying specifics.

Phase 1: Discovery

#!/bin/bash

echo "=== Pytest Configuration ==="
if [ -f "pytest.ini" ]; then
    cat pytest.ini | head -20
elif [ -f "pyproject.toml" ]; then
    grep -A 20 "\[tool.pytest" pyproject.toml | head -20
elif [ -f "setup.cfg" ]; then
    grep -A 20 "\[tool:pytest\]" setup.cfg | head -20
fi

echo ""
echo "=== Pytest Version ==="
python -m pytest --version 2>/dev/null

echo ""
echo "=== Test File Summary ==="
find . -name "test_*.py" -o -name "*_test.py" 2>/dev/null | grep -v __pycache__ | wc -l | xargs echo "Total test files:"
find . -name "test_*.py" 2>/dev/null | grep -v __pycache__ | head -15

echo ""
echo "=== Conftest Files ==="
find . -name "conftest.py" 2>/dev/null | grep -v __pycache__ | head -10

echo ""
echo "=== Installed Plugins ==="
python -m pytest --co -q 2>&1 | tail -3

Phase 2: Analysis

#!/bin/bash

echo "=== Test Collection (dry run) ==="
python -m pytest --collect-only -q 2>/dev/null | tail -5

echo ""
echo "=== Markers ==="
python -m pytest --markers 2>/dev/null | head -15

echo ""
echo "=== Coverage Summary ==="
if [ -f "htmlcov/status.json" ]; then
    cat htmlcov/status.json | python -c "
import json, sys
d = json.load(sys.stdin)
print(f'Lines: {d.get(\"totals\",{}).get(\"percent_covered_display\",\"N/A\")}%')
" 2>/dev/null
elif [ -f ".coverage" ]; then
    python -m coverage report --show-missing 2>/dev/null | tail -10
fi

echo ""
echo "=== Last JUnit Report ==="
REPORT=$(find . -name "report.xml" -o -name "junit.xml" -o -name "test-results.xml" 2>/dev/null | head -1)
if [ -n "$REPORT" ]; then
    pytest_results "$REPORT" | python -m json.tool
fi

Output Rules

• TOKEN EFFICIENCY: Target 50 lines per output
• Use --tb=short or --tb=line for concise failure output
• Parse JUnit XML or JSON reports rather than raw console output

Anti-Hallucination Rules

• NEVER guess test function names — always discover via --collect-only
• NEVER fabricate coverage numbers — parse actual .coverage or reports
• NEVER assume conftest fixtures — check conftest.py files first

Safety Rules

• NEVER modify test files without explicit user confirmation
• NEVER delete .coverage or htmlcov without user approval
• Be cautious with --lf (last failed) — it depends on cache state

Output Format

Present results as a structured report:

Managing Pytest Report
══════════════════════
Resources discovered: [count]

Resource       Status    Key Metric    Issues
──────────────────────────────────────────────
[name]         [ok/warn] [value]       [findings]

Summary: [total] resources | [ok] healthy | [warn] warnings | [crit] critical
Action Items: [list of prioritized findings]

Target ≤50 lines of output. Use tables for multi-resource comparisons.

Counter-Rationalizations

Shortcut	Counter	Why
"I'll skip discovery and check known resources"	Always run Phase 1 discovery first	Resource names change, new resources appear — assumed names cause errors
"The user only asked for a quick check"	Follow the full discovery → analysis flow	Quick checks miss critical issues; structured analysis catches silent failures
"Default configuration is probably fine"	Audit configuration explicitly	Defaults often leave logging, security, and optimization features disabled
"Metrics aren't needed for this"	Always check relevant metrics when available	API/CLI responses show current state; metrics reveal trends and intermittent issues
"I don't have access to that"	Try the command and report the actual error	Assumed permission failures prevent useful investigation; actual errors are informative