oncokb-datastar-review
Review Datastar code for Tao-aligned architecture, correct attributes/actions usage, and maintainable server-driven UI patterns.
SKILL.mdskills/oncokb-datastar-review/SKILL.md
GitHubScope
- Use when a change adds or modifies Datastar attributes, actions, signals, or Datastar-driven SSE flows.
- Focus on review findings, not feature implementation.
Workflow
Detect Datastar surface area in the change.
- Identify changed files using
data-*Datastar attributes,@action(...)expressions, SSE patch events, and Datastar-specific JS/plugin code. - Completion criterion: Datastar-affected files and interaction paths are explicitly listed.
- Identify changed files using
Check Tao principles first.
- Apply
reference/the-tao-of-datastar.mdas architecture-level guardrails. - Verify backend-as-source-of-truth, minimal frontend state, and server-driven patching.
- Completion criterion: each applicable Tao principle has a pass/fail note with evidence.
- Apply
Check attribute/action correctness.
- Verify attribute semantics, modifiers, casing, and order-sensitive patterns.
- Verify backend action method/option usage (
@get,@post, retries, cancellation, signal filtering). - Completion criterion: no undocumented or incorrect Datastar attribute/action usage remains unreported.
Check interaction quality and reliability.
- Flag optimistic updates that claim success before backend confirmation.
- Confirm loading states/indicators for long-running operations.
- Confirm morph/patch behavior preserves intended state (
data-ignore-morph,data-preserve-attrwhen required). - Completion criterion: UX and reliability risks are either accepted explicitly or documented as findings.
Report actionable review findings.
- Separate blockers from non-blocking improvements.
- Include file paths, principle/rule violated, and a concrete fix direction.
- Completion criterion: author can apply each finding without additional discovery.
Output format
Blockers: incorrect Datastar semantics, architecture violations, or high-risk UX/reliability defects.Non-blocking: maintainability and clarity improvements.Checked: Tao principles reviewed, attribute/action correctness reviewed, and reliability checks completed.
Quality bar
- Prefer Tao-aligned server-driven patterns over frontend state accumulation.
- Do not approve speculative/optimistic-success UI without explicit product agreement.
- Keep findings concrete and tied to changed code.
Reference
reference/the-tao-of-datastar.md../oncokb-datastar/reference/datastar-attributes.md../oncokb-datastar/reference/datastar-actions.md../oncokb-datastar/reference/examples/index.md