extract-foundations-from-screens

Act as a design system extractor that turns a cross-screen inventory of raw values into canonical Foundations docs. Goal: Read a set of `SCREEN_SPEC.md` files, union the raw values they carry (hex colors, typography tuples, spacing pixels, radii pixels), bucket each value into a role token per the per-foundation rules below, and write or update `foundations/color.md`, `foundations/typography.md`, `foundations/spacing.md`, `foundations/radii.md` against the existing FOUNDATION templates. The operation is idempotent: re-running on a superset of screens only adds new tokens; existing role tokens are never overwritten; conflicts land in a `## Review queue` section instead of being silently resolved. Mandatory context bootstrap (before any output): <!-- shared:mandatory-context-bootstrap --> - Read these sections in `WORKFLOW_OPERATING_SYSTEM.md` first: - `## LLM execution contract` - `## Editor mode policy` (mode definitions only; the tool mapping table is lazy-loaded in `wos/editor-mode-mappings.md` and needed only for non-Claude-Code tools) - `## Global output contract` (including **Adaptive handoff** and **Mode selection rule**) - `## Cross-cutting workflow guardrails` - **Bootstrap tiers (ADR-0025):** the light-weight commands (`branch-commit`, `what-next`, `where-we-at`, `slice-closure`, `compact-task-memory`) may skip `## Editor mode policy` good-fits lists and `## Cross-cutting workflow guardrails` sequencing heuristics, reading only the mode definitions and the ...