Implement pagination for search results and memory listings

[BunsDev]

Problem

Queries currently load results with high arbitrary limits and no proper pagination. As the database grows, this will cause:

• Excessive memory usage on the server
• Slow page loads for users
• Potential browser performance issues rendering large lists

Affected areas: memory entry listings, trace search, artifact explorer, investigation results.

Proposed Solution

Implement cursor-based pagination:

1. API layer — Accept cursor and limit query parameters in search/list endpoints
2. DB layer — Use keyset pagination (WHERE id > ? ORDER BY id LIMIT ?) for consistent performance
3. UI layer — Add "Load more" or infinite scroll to list views
4. Response format — Return { data: [...], nextCursor: string | null, hasMore: boolean }

Acceptance Criteria

• All list/search endpoints support cursor and limit parameters
• Default page size is reasonable (e.g., 25-50 items)
• UI shows pagination controls or infinite scroll
• Performance is consistent regardless of total dataset size
• Backward compatible — omitting pagination params returns first page

🤖 Generated with Claude Code