run-validation

name	run-validation
description	Reusable validation procedure for build, test, lint, architecture, and security checks. Use when verifying changes before review completion or final handoff.

Skill: Run Validation

Reusable procedure for validating changes before marking work as complete.

When to Use

• After implementation, before marking a task done
• After fixing review or security findings
• Before the orchestrator declares completion
• Whenever targeted verification is needed

Procedure

1. Build Check

pnpm run build

• Must exit 0
• If it fails, read the error output and fix the specific issue

2. Test Check

pnpm run test

• Must exit 0
• If specific packages changed, use a targeted command where possible

3. Lint Check

pnpm run lint

• Must exit 0
• Auto-fix is available through pnpm run format

4. Architecture Tests

pnpm run test:arch

• Preferred repo-wide command from the root package.json; use when architecture boundaries changed or when validating a PR before handoff

Focused package commands:

pnpm --filter @ocom-verification/archunit-tests test:arch
pnpm --filter @cellix/archunit-tests test:arch

• Use when domain, GraphQL, persistence, or architecture boundaries changed

5. Security Scan

pnpm run snyk

• Use for dependency or security-relevant changes when the environment supports it

6. Full Verification

pnpm run verify

• Use before final completion when broad verification is appropriate

Validation Matrix

Change Type	Build	Test	Lint	Arch	Snyk	Acceptance
Domain logic	yes	yes	yes	yes	-	domain
GraphQL schema or resolver	yes	yes	yes	optional	-	graphql
Persistence layer	yes	yes	yes	optional	-	optional
UI components	yes	yes	yes	-	-	ui
Dependencies	yes	yes	yes	-	yes	-
Auth or security code	yes	yes	yes	optional	yes	yes

Failure Recovery

1. Read the error output carefully
2. Identify the root cause rather than the surface symptom
3. Fix the specific issue
4. Re-run the failing validation
5. Re-run earlier checks if the fix could have regressed them