Write an eval suite for the pf-design-mode skill (design-guide plugin). This skill creates and edits Figma design files using PF-approved component libraries. It requires the Figma MCP.
Discriminating signal: PF-approved component library selection, correct Figma component structure, auto-layout patterns per PF standards. Baseline Claude may suggest generic Figma operations that don't use PF component libraries.
MCP dependency: Figma MCP required. Mock frame/component data as eval input. Tag Figma-dependent cases with requires_mcp: figma for gated CI execution.
Acceptance criteria:
-
eval/pf-design-mode/eval.yaml exists following the pf-unit-test-generator template
-
5+ test cases covering explicit invocation, implicit/contextual prompts, and at least 1 negative control (e.g., non-PF design request)
-
Inline Python judges test discriminating behavior (PF component library usage, not general Figma operations)
-
With-skill/without-skill A/B delta documented
-
All judges pass at defined thresholds
Jira Issue: PF-4342
Write an eval suite for the
pf-design-modeskill (design-guideplugin). This skill creates and edits Figma design files using PF-approved component libraries. It requires the Figma MCP.Discriminating signal: PF-approved component library selection, correct Figma component structure, auto-layout patterns per PF standards. Baseline Claude may suggest generic Figma operations that don't use PF component libraries.
MCP dependency: Figma MCP required. Mock frame/component data as eval input. Tag Figma-dependent cases with
requires_mcp: figmafor gated CI execution.Acceptance criteria:
eval/pf-design-mode/eval.yamlexists following the pf-unit-test-generator template5+ test cases covering explicit invocation, implicit/contextual prompts, and at least 1 negative control (e.g., non-PF design request)
Inline Python judges test discriminating behavior (PF component library usage, not general Figma operations)
With-skill/without-skill A/B delta documented
All judges pass at defined thresholds
Jira Issue: PF-4342