Web Interface Guidelines Reviewer
Automatically review UI code against web interface guidelines for accessibility and best practices.
The skill mandates fetching latest guidelines from an external URL in each review, potentially introducing untrusted content, and lacks user confirmation or permission isolation. It requires read and network permissions without explicit least-privilege or sensitive-data handling. Score 8, deduct 17: no data-flow disclosure, no rollback/isolation, unverifiable source (externally fetched rules not auditable).
Low self-consistency: guideline fetching steps are vague (e.g., 'fetch latest guidelines'), and no error handling for network unavailability or missing files. Static review cannot execute key paths, and no test evidence. Score 2, deduct 8: no reproducible tests or failure diagnostics, unclear failure feedback.
Clear scenario: review UI code, but trigger conditions may mismatch actual capability (requires external network and specific tools). Non-fit ranges not declared, and no consideration for mainland-China reachability (relies on raw.githubusercontent.com, likely unreachable). Score 5, deduct 10: boundaries undefined, insufficient environment-fit evidence.
Minimal documentation: no install/dependency notes, changelog, maintenance responsibility, or known limitations. Metadata provides version and author, but lacks change history and provenance verification. Score 3, deduct 12: lacks layered docs, examples, and troubleshooting.
Cannot verify task completion: depends on externally fetched guideline format, and output format not specified in this file. Static review cannot test, and no examples provided. Score 1, deduct 14: unverifiable value claim, insufficient direct-usability evidence.
No tests or third-party execution evidence. All key claims (e.g., fetch guidelines, apply rules) are author assertions, not independently verifiable. Score 1, deduct 9: lacks auditable test results or cross-validation.
- The skill relies on fetching external content from raw.githubusercontent.com, which may be unreachable in mainland China; consider mirroring or bundling rules.
- No user confirmation mechanism; executing may automatically read files and make network requests; use with caution.
- Externally fetched guideline content is not auditable and may contain malicious or inaccurate instructions; verify source.
What does this skill do, and when should you use it?
This skill reviews UI code for compliance with the Web Interface Guidelines. It fetches the latest guidelines from a specified URL, reads user-specified files, applies the rules to the code, and outputs findings in a terse `file:line` format. It is part of the Multica platform's reusable skills collection, designed to support code review and design audit workflows.
Fetches the latest Web Interface Guidelines from a source URL (vercel-labs/web-interface-guidelines command.md); reads specified files or directory patterns; checks each file against all rules in the fetched guidelines; outputs findings in file:line format; prompts the user for files if no argument is provided.
- A frontend developer wants to quickly check their current UI code against the latest web accessibility best practices.
- A design system maintainer wants to review component library code for compliance before merging.
- A QA engineer performs a design audit on pages before release to catch potential accessibility issues.
- A tech lead automates design guideline checks as part of the code review process.
- A developer receives a request to 'review my UI' or 'audit design' and uses this skill to provide structured feedback quickly.
What are this skill's strengths and limitations?
- Fetches the latest guidelines automatically, ensuring reviews are based on up-to-date rules.
- Terse output format is easy to integrate into CI/CD or code review tools.
- Flexible file or pattern specification for the review scope.
- As part of Multica's skill set, it can work together with other skills.
- Depends on an external URL; if network is unavailable or the URL breaks, the skill fails.
- No test suite documented, error handling behavior unknown.
- Only outputs `file:line` findings, may lack detailed remediation suggestions.
- No explicit support for specific frameworks or languages; may not work on non-web code.
- Attribution to vercel as author is present, but maintenance status is unclear.
How do you install this skill?
The skill is located in the Multica repository under .agents/skills/web-design-guidelines/. After installing the Multica platform, the skill is included by default as part of the skill collection. Specific standalone installation steps are not documented in the source.
How do you use this skill?
Trigger the skill by providing a file or pattern to review, e.g., 'review src/components/Button.tsx'. The skill will fetch the latest guidelines, read the specified files, apply the rules, and output findings in file:line format. If no files are specified, it asks the user which files to review.