name	ops-hub-validation
description	Use to select and run Personal Ops Hub validation commands for Kotlin/Ktor, dashboard JavaScript/CSS/HTML, shell scripts, API smoke checks, and browser screenshots.

Ops Hub Validation

Baseline Checks

Run these for most code changes:

./gradlew test processResources
node --check src/main/resources/dashboard/dashboard.js
git diff --check

Run shell syntax checks when scripts or deployment docs change:

bash -n scripts/install-systemd.sh scripts/update-server.sh scripts/report-backup.sh

Run harness structural validation when harness files change:

python3 ~/.codex/skills/codex-harness/scripts/validate_harness.py .codex/harnesses/personal-ops-hub

Targeted Checks

Backend route/service/config changes: ./gradlew test, plus API smoke if the route is not fully covered.
Database or retention changes: add tests for old/new rows, protected states, and migration behavior.
Dashboard JS changes: node --check, ./gradlew processResources, and browser smoke.
Dashboard CSS/layout changes: browser screenshot at desktop and relevant mobile viewport.
Systemd/Docker/log/action changes: fake binary smoke tests where possible, and tests for allowlist refusal paths.
Install/update scripts: bash -n, README/runbook review, and public placeholder scan.
Docs-only changes: structural validation or link/path sanity as relevant; code tests may be skipped with a reason.

Browser Smoke Pattern

Use the Browser plugin or Playwright for local dashboard targets. Save screenshots under output/playwright/ with descriptive names.

Confirm the specific behavior changed, not just that the page loads. Examples:

Theme: dataset.theme, localStorage, and selected segmented control state.
Mobile: scrollWidth <= clientWidth for nav, or expected card/list display for inventory rows.
Actions/logs: expected buttons exist only for allowed rows and refused actions show readable feedback.
Event UX: labels, selected state, detail panel copy, and button spacing.

Evidence Format

Record:

Command.
Pass/fail.
Important output line or screenshot path.
Skipped checks and why they were not applicable.

Ops Hub Validation

Baseline Checks

Run these for most code changes:

./gradlew test processResources node --check src/main/resources/dashboard/dashboard.js git diff --check

Run shell syntax checks when scripts or deployment docs change:

bash -n scripts/install-systemd.sh scripts/update-server.sh scripts/report-backup.sh

Run harness structural validation when harness files change:

python3 ~/.codex/skills/codex-harness/scripts/validate_harness.py .codex/harnesses/personal-ops-hub

Targeted Checks

Backend route/service/config changes: ./gradlew test, plus API smoke if the route is not fully covered.

Database or retention changes: add tests for old/new rows, protected states, and migration behavior.

Dashboard JS changes: node --check, ./gradlew processResources, and browser smoke.

Dashboard CSS/layout changes: browser screenshot at desktop and relevant mobile viewport.

Systemd/Docker/log/action changes: fake binary smoke tests where possible, and tests for allowlist refusal paths.

Install/update scripts: bash -n, README/runbook review, and public placeholder scan.

Docs-only changes: structural validation or link/path sanity as relevant; code tests may be skipped with a reason.

Browser Smoke Pattern

Use the Browser plugin or Playwright for local dashboard targets. Save screenshots under output/playwright/ with descriptive names.

Confirm the specific behavior changed, not just that the page loads. Examples:

Theme: dataset.theme, localStorage, and selected segmented control state.

Mobile: scrollWidth <= clientWidth for nav, or expected card/list display for inventory rows.

Actions/logs: expected buttons exist only for allowed rows and refused actions show readable feedback.

Event UX: labels, selected state, detail panel copy, and button spacing.

Evidence Format

Record:

Command.

Pass/fail.

Important output line or screenshot path.

Skipped checks and why they were not applicable.

ops-hub-validation