# Testing And Validation

> PetSwipe validation workflow for choosing the smallest reliable checks for frontend, backend, infra, docs, and mixed changes. Use when a task needs a concrete validation plan or when the repo's pre-existing lint or type issues make broad validation noisy.

- Skill: `hoangsonww/testing-and-validation` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add hoangsonww/testing-and-validation`
- Raw SKILL.md: https://api.skillmd.com/api/skills/hoangsonww/testing-and-validation/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Web & Frontend
- Author: hoangsonww (https://skillmd.com/u/hoangsonww)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/hoangsonww/testing-and-validation

---


Use this skill when validation strategy matters as much as the code change.

## Workflow

1. Identify the touched area and its real blast radius.
2. Choose the smallest reliable checks that validate that area.
3. Prefer targeted checks before broad repo-wide checks.
4. Report clearly what each command validates.
5. Separate new failures from pre-existing repo debt.

## PetSwipe Guidance

- Frontend: prefer targeted `npx eslint` over broad linting when possible.
- Backend: prefer `npm run build` plus targeted tests.
- Infra: prefer render and preflight commands.
- Docs: validate by re-reading commands, paths, and claims against the repo.

If needed, spawn `test_guard` to recommend the validation set before editing.

Read `references/validation-playbook.md`.