# Validate > Manually test your work against the locally running app by making real API or HTTP requests. Use when this capability is needed. - Skill: `tomevault-io/validate-2` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/validate-2` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/validate-2/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Integrations & APIs - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/validate-2 --- Validate the changes you've made by testing them against the locally running application: 1. Figure out what you changed and what needs testing — review recent diffs or your current task context. 2. Determine how to reach the app (check for running servers, common ports, env files, docker-compose, etc.). 3. Make real requests using `curl`, `httpie`, or similar tools to exercise the affected endpoints or functionality. 4. Verify the responses are correct — check status codes, response bodies, error cases, and edge cases. 5. If something is broken, fix it and re-test. 6. Report a summary of what you tested and the results. Be thorough: don't just hit the happy path. Test error cases, missing fields, invalid input, and boundary conditions where relevant. --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/mjwestcott) — claim your Tome and manage your conversions.