chore: release main - #3

* feat(web): cost affordance at the point of paid action (P1.6)

UX audit P1.6 (CPO priority #3 — tokens are the community's #1 measurable
pain; surprise-cost = churn). The paid actions offered no cost cue at the
trigger. Surfaces the existing CostBadge (spend = brand orange, not red — a
chosen evaluation is the valuable action, not a hazard) on the roomy
single-action surfaces, mirroring Explore's token-honesty:
- GeneratePdfButton (component-level → shows in the report viewer): a 'Uses
  tokens' badge beside 'Generate tailored CV' in its not-yet-generated state;
- QuickEvaluate (home): a one-line cue + badge under the paste field —
  'Evaluation runs on your own AI — your key, your machine.'
The dense pipeline-row Evaluate button is deliberately left for P1.4 (row
hierarchy), where a per-row badge would fight the noise-reduction goal.
Verified E2E: report santifer#87 (no PDF) shows the badge; home in-between shows the
cue; typecheck clean.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* polish(web): cost 'spend' badge is muted, not brand-orange (design-review)

career-ops-ux system correction on santifer#1536: orange is reserved for the primary
action (incl. 'Run your first FREE scan'), so an orange 'Uses tokens' badge
collides semantically ('go/free' vs 'costs') AND — now the button is neutral —
makes the warning shout louder than the action. Aligns to the color-semantics
lock (-> DESIGN_SYSTEM.md): GREEN = free/celebrate, NEUTRAL/muted (+ coin) =
spend/inform, ORANGE = primary action (reserved). CostBadge is the shared
primitive, so this one change propagates to every cost cue — Explore, the
report's Generate-CV, and P1.4's inbox legend — as a single pattern. Muted
AA-verified: 5.68 light / 8.34 dark.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>