SKILL PLAN · DETERMINISTIC
A small plan for a real job
Start with the minimum useful stack. Add nothing unless it has a clear responsibility in the run.
- Task
- UI review
- Agent
- Codex
- Priority
- Balanced
Rules + reviewed metadata only. No LLM. Same inputs produce the same plan.
01Core execution
Native · Codex
web-design-guidelines
Runs the structured interface review. Open profile and install guide ↗Review permissions, limitations, and evidence
- Agent fit
- Native · Codex
- Main limitation
- It depends on network access and a live guideline document whose contents can change over time.
- First risk
- Rule drift or context-free findings · Medium
- Source
- Pinned source · b8caa260
02Guardrail
Portable · Codex
context-map
Keeps the review tied to explicit rules and project context. Open profile and install guide ↗Review permissions, limitations, and evidence
- Agent fit
- Portable · Codex
- Main limitation
- The workflow is intentionally brief and may miss dynamic dependencies or runtime ownership.
- First risk
- Incomplete map treated as full scope · Medium
- Source
- Pinned source · 318066d2
03Evidence
Portable · Codex
webapp-testing
Captures comparable visual evidence for the final decision. Open profile and install guide ↗Review permissions, limitations, and evidence
- Agent fit
- Portable · Codex
- Main limitation
- It assumes Python Playwright and a runnable local application are available.
- First risk
- Browser actions against the wrong target · High
- Source
- Pinned source · 0a64e398
Combined permission budget
This is the union of declared permissions across the plan. Review each package before installation.
Read frontend filesRead filesBrowser controlRun commandsWrite screenshots
Finish with evidence
Turn a vague design review into a short, evidence-backed fix list.
Review before editing; every suggested fix must point to visible evidence or source lines.
- 3-step acceptance check
- Review the primary page at three representative widths.
- Verify focus order, labels, contrast, and the main interaction.
- Keep only findings that include reproduction evidence.
- Stop conditions
- The review claims standards compliance without testing.
- Screenshots are used as proof of behavior.
- Suggested changes conflict with the product's established design system.