perf: optimize API fetching with caching and debouncing (#11)

[Subramaniyajothi6]

📌 Description

Optimizes client-side API fetching by adding a lightweight in-memory GET cache and debouncing search inputs. This cuts redundant network requests during page navigation and rapid typing, with no external libraries added.

🔗 Related Issue

Closes #11

🛠 Changes Made

• Add src/utils/apiCache.js — in-memory GET cache (60s TTL, prefix-based invalidate, clearCache, and in-flight request de-duplication so concurrent identical requests share one call).
• Add src/hooks/useDebounce.js — generic value-debounce hook.
• Route reads through the cache: tasks, routines, journal (list / by-date / analytics), and analytics.
• Invalidate the relevant cache keys on every mutation (task add/update/delete/bulk, routine create/duplicate/delete, journal save/delete) so stale data is never served; clear the cache on logout.
• Debounce the Daily Journal search (server-side fetch that previously fired one request per keystroke) and the Tasks search (client-side filter recompute).

📸 Screenshots (if applicable)

N/A — behavior-only optimization, no UI changes. Fewer/de-duplicated requests are visible in the browser Network tab.

✅ Checklist

• Code runs locally
• Followed project structure
• No console errors
• Properly tested changes
• Linked the issue

🚀 Notes for Reviewers

• Inputs stay fully responsive; only the value the filter/fetch reads is debounced.
• Mutations refetch after invalidation, so the cache never serves stale data — savings come from cross-page navigation reuse and request de-duplication.
• The Tasks search is client-side, so debouncing it reduces re-render/recompute rather than API calls; the Daily Journal search is server-side and now sends far fewer requests.
• npm run lint and npm run build pass in frontend/.

@aryandas2911 Please review this pr