Frontend Review
Defines the preconditions and environmental requirements for running frontend design review and accessibility audit methods.…
$ prime install @community/scope-frontend-review Projection
Always in _index.xml · the agent never has to ask for this.
FrontendReview [scope] v1.0.0
Defines the preconditions and environmental requirements for running frontend design review and accessibility audit methods. Agents must verify these conditions before executing @community/method-a11y-audit or @community/method-heuristic-review.
Loaded when retrieval picks the atom as adjacent / supporting.
FrontendReview [scope] v1.0.0
Defines the preconditions and environmental requirements for running frontend design review and accessibility audit methods. Agents must verify these conditions before executing @community/method-a11y-audit or @community/method-heuristic-review.
Label
Frontend Review Execution Environment
Preconditions
- Id: artifact-available
- Check: artifact input is a valid URL (fetchable), local file path (readable), base64 image, or raw HTML string
- On Fail: Request the artifact from the user before proceeding.
- Id: axe-core-available
- Check: axe-core@^4.9 is importable in the current Node.js environment OR the artifact is a URL and Playwright can be used for browser-based analysis
- On Fail: Warn that automated axe-core analysis will be skipped; proceed with manual + visual analysis only.
- Id: sufficient-context
- Check: remaining context budget > 2000 tokens
- On Fail: Compact context before starting a full audit. Load only summary.md chunks for referenced atoms.
Environment
- Runtime: Node.js >= 20 OR Bun >= 1.0
- Browser: Chromium (via Playwright) for live URL analysis
- Required Packages:
- axe-core@^4.9
- lighthouse@^12
- Optional Packages:
- @playwright/test@^1.40
Token Budget
- Method A11y Audit: ~800 tokens (summary level atoms)
- Method Heuristic Review: ~600 tokens (summary level atoms)
- Full Deep Review: ~2000 tokens (core level atoms + violation details)
Loaded when retrieval picks the atom as a focal / direct hit.
FrontendReview [scope] v1.0.0
Defines the preconditions and environmental requirements for running frontend design review and accessibility audit methods. Agents must verify these conditions before executing @community/method-a11y-audit or @community/method-heuristic-review.
Label
Frontend Review Execution Environment
Preconditions
- Id: artifact-available
- Check: artifact input is a valid URL (fetchable), local file path (readable), base64 image, or raw HTML string
- On Fail: Request the artifact from the user before proceeding.
- Id: axe-core-available
- Check: axe-core@^4.9 is importable in the current Node.js environment OR the artifact is a URL and Playwright can be used for browser-based analysis
- On Fail: Warn that automated axe-core analysis will be skipped; proceed with manual + visual analysis only.
- Id: sufficient-context
- Check: remaining context budget > 2000 tokens
- On Fail: Compact context before starting a full audit. Load only summary.md chunks for referenced atoms.
Environment
- Runtime: Node.js >= 20 OR Bun >= 1.0
- Browser: Chromium (via Playwright) for live URL analysis
- Required Packages:
- axe-core@^4.9
- lighthouse@^12
- Optional Packages:
- @playwright/test@^1.40
Token Budget
- Method A11y Audit: ~800 tokens (summary level atoms)
- Method Heuristic Review: ~600 tokens (summary level atoms)
- Full Deep Review: ~2000 tokens (core level atoms + violation details)
Source
prime-system/examples/frontend-design/primes/compiled/@community/scope-frontend-review/atom.yaml