feat: cache repeated searches for 60 seconds

[Heazzy500]

Summary

• add a bounded 100-entry LRU cache with a 60-second TTL
• key entries by query, result count, and freshness
• count cache hits while returning cached: true
• only cache successful Serper responses after payment middleware has accepted the request

Closes #21

[vercel]

@zeemscript is attempting to deploy a commit to the Emmanuel's projects Team on Vercel.

A member of the Team first needs to authorize it.

[drips-wave]

@Heazzy500 Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits

[AbuJulaybeeb]

resolve conflicts

[Heazzy500]

@Emmy123222 The merge conflicts for this PR have been resolved against the latest main and pushed to Heazzy500:fix/issue-21. Please re-run/inspect CI when available.

[AbuJulaybeeb]

@Emmy123222 The merge conflicts for this PR have been resolved against the latest main and pushed to Heazzy500:fix/issue-21. Please re-run/inspect CI when available.

@Heazzy500 fetch upstream main, rebase upstream/main, then u push

[Heazzy500]

@Emmy123222 Merge conflicts resolved for this PR against latest main. Rebased fix/issue-21 onto upstream/main and pushed — new head: 5cbd796. Conflict files: package.json, package-lock.json, server/index.ts (plus src/types/index.ts change to type the cached field on SearchResponse). Validation: git diff --check clean, working tree clean, remote head verified; node_modules not present so full typecheck not run. Please review / re-run CI when convenient. Thanks!

[Heazzy500]

@Emmy123222 @Heazzy500 PR #71 was rebased onto the current upstream main. The cache implementation is now combined with the current search route, preserving request reconciliation and validation while adding bounded TTL caching. The repaired branch is pushed at commit 9e273b3 and ready for CI/review.