playbooks

home / skills / levnikolaevich / claude-code-skills / ln-514-regression-checker

ln-514-regression-checker skill

/ln-514-regression-checker

This skill automatically runs the existing test suite to detect regressions and reports pass/fail without altering task status for quality gates.

npx playbooks add skill levnikolaevich/claude-code-skills --skill ln-514-regression-checker

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

Files (3)
SKILL.md
1.9 KB
---
name: ln-514-regression-checker
description: Worker that runs existing tests to catch regressions. Auto-detects framework, reports pass/fail. No status changes or task creation.
---

> **Paths:** File paths (`shared/`, `references/`, `../ln-*`) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root.

# Regression Checker

Runs the existing test suite to ensure no regressions after implementation changes.

## Purpose & Scope
- Detect test framework (pytest/jest/vitest/go test/etc.) and test dirs.
- Execute full suite; capture results for Story quality gate.
- Return PASS/FAIL with counts/log excerpts; never modifies Linear or kanban.

## When to Use
- **Invoked by ln-510-quality-coordinator** Pass 1 (after ln-511 and ln-513)
- Code quality check passed

## Workflow (concise)
1) Auto-discover test framework per `shared/references/ci_tool_detection.md` Command Registry (Test Frameworks section).
2) **Read `docs/project/runbook.md`** — get exact test commands, Docker setup, environment variables. Runbook commands take priority over auto-detection (per ci_tool_detection.md Discovery Hierarchy).
3) Build appropriate test command; run with timeout (5min per ci_tool_detection.md); capture stdout/stderr.
4) Parse results: passed/failed counts; key failing tests.
5) Output verdict JSON (PASS or FAIL + failures list) and add Linear comment.

## Critical Rules
- No selective test runs; run full suite.
- Do not fix tests or change status; only report.
- Language preservation in comment (EN/RU).

## Definition of Done
- Framework detected; command executed.
- Results parsed; verdict produced with failing tests (if any).
- Linear comment posted with summary.

## Reference Files
- Risk-based limits used downstream: `../shared/references/risk_based_testing_guide.md`
- **CI tool detection:** `shared/references/ci_tool_detection.md`

---
**Version:** 3.1.0
**Last Updated:** 2026-01-09

Overview

This skill runs the project’s full test suite to detect regressions after code changes. It auto-detects the test framework, respects the project runbook for exact commands, and returns a clear PASS/FAIL verdict with counts and failure excerpts. It reports findings via a Linear comment but never alters issue status or creates tasks.

How this skill works

I discover the test framework and test directories using the CI tool detection rules, then read the project runbook to prefer explicit test commands and environment setup. I execute the full test suite with a strict timeout, capture stdout/stderr, parse passed/failed counts and identify top failing tests, and produce a JSON verdict and Linear comment summarizing results.

When to use it

  • After code changes that require validation of existing functionality
  • Invoked as Pass 1 by the quality coordinator after earlier quality checks complete
  • When you need an automated, non-invasive regression check before release
  • When the project runbook or CI detection indicates available test suites

Best practices

  • Ensure docs/project/runbook.md contains canonical test commands and environment notes
  • Keep tests and test directories discoverable by standard frameworks (pytest, jest, vitest, go test, etc.)
  • Do not rely on this skill to triage or fix failures; treat output as a signal for follow-up work
  • Provide container/Docker setup in the runbook if tests require special environments
  • Allow enough CI timeout for slow suites; default is 5 minutes unless runbook specifies otherwise

Example use cases

  • Run a full suite after a feature branch merge to catch regressions before QA
  • Automate regression checks in a CI pipeline invoked by the quality coordinator
  • Produce a concise pass/fail summary and failure excerpts for story-level quality gates
  • Verify that test infrastructure and runbook commands remain accurate after refactors

FAQ

Does this skill change issue status or create tasks in Linear?

No. It only posts a comment with the summary and verdict; it never modifies statuses or creates tasks.

What if the runbook and auto-detection disagree on the test command?

The runbook takes precedence. If runbook commands are missing, the skill falls back to auto-discovered commands per CI detection rules.