Skip to main content

​
Summary

Two-part session: (1) visual polish of the smo1.io landing page, (2) full scan of all GitHub issues across the 6-repo smo1-io organisation to align tracker state with codebase reality.

​
What changed

​
Landing page (shipped)

The whiskers-landing site had broken dark mode across all card and button components. Root cause: the CSS file was missing design system color tokens — only cat-themed status colors existed. Added 16 tokens, fixed navbar overlap, improved hero conversion layout, and fixed testimonials grid. Pushed to main.

​
Issue alignment (7 updates)

Scanned 24 issues across whiskers-landing, meow-web, purr-api, zoomies-edge, catnip-packages, and litter-box. Found 6 stale or inaccurate issues. Created 1 new critical bug issue. Posted corrective comments on 6 existing issues. Updated the production readiness EPIC with corrected statuses.

​
Key findings with business impact

​
Shared types package is broken (P0)

The @smo1/types package — used by every frontend service — has a duplicate type definition producing a broken merged interface. This blocks the UI component library migration (catnip-packages#2) and dashboard component sharing. Filed as catnip-packages#10. Estimated 2-3 hours to fix.

​
KV cache invalidation is further along than tracked

The issue tracker said the Cloudflare KV sync service needed to be built from scratch. In reality, the full implementation already exists in purr-api with sync, delete, and bulk operations. What actually remains is test coverage and wiring verification — roughly 2-3 hours, not the 8-12 originally estimated.

​
Test infrastructure has a hidden blocker

The purr-api Go backend has an import cycle in its repository test files that silently prevents those tests from running. This means the “50% coverage” estimate in the test tracking issue is unverifiable for the repository layer.

​
Operational takeaways

  1. Issue tracker drift is real. After 6 days without a full scan, 5 of 11 open issues had stale or incorrect information. Regular cross-repo scans (weekly or per-sprint) should be standard practice.
  2. CSS token completeness matters. A missing design system foundation caused cascading visual bugs across every component. When adopting shadcn/ui or similar token-based systems, validate the full token set before building components on top.
  3. Test coverage numbers need verification. Reported coverage percentages are only meaningful if the test runner can actually execute all test files. Silent failures (import cycles, missing containers) create false confidence.

​
Action items

  • Fix @smo1/types duplicate Link interface and missing SubscriptionTier (catnip-packages#10) — blocks UI package migration
  • Add kv_sync_service tests in purr-api — unblocks closing zoomies-edge#3
  • Fix import cycle in purr-api repository tests — unblocks accurate coverage measurement
  • Schedule weekly cross-repo issue scan to prevent tracker drift
  • Validate shadcn token completeness in meow-web globals.css (same pattern may exist)