Implement client-side Stockfish/Maia analysis caching with auto-save and load functionality

[Copilot]

This PR implements automatic caching of client-side Stockfish and Maia analysis to prevent loss when users navigate away from the analysis page.

Problem

Users were losing all their analysis progress whenever they left the analysis page, requiring them to re-analyze positions they had already studied.

Solution

Added comprehensive auto-save and load functionality:

Auto-Save Mechanism

• Automatic saves every 10 seconds via background timer
• Immediate saves when new analysis is detected (debounced)
• Final save on component unmount or game change
• Smart filtering - only saves meaningful analysis (depth ≥ 12 or Maia data)
• Cache optimization - avoids redundant saves using cache key comparison

Load Mechanism

• Automatic loading when opening any supported game type
• FEN verification ensures stored analysis applies to correct positions
• Protection against overwriting deeper analysis with shallower stored data
• Background operation - failures don't interrupt user experience

Implementation Details

New API functions (src/api/analysis/analysis.ts):

storeEngineAnalysis(gameId: string, analysisData: EngineAnalysisData)
getEngineAnalysis(gameId: string): Promise<EngineAnalysisData | null>

Analysis storage utilities (src/lib/analysisStorage.ts):

• collectEngineAnalysisData() - extracts analysis from GameTree nodes
• applyEngineAnalysisData() - applies stored analysis back to tree
• generateAnalysisCacheKey() - prevents unnecessary saves

Data format matches backend specification:

{
  "positions": [
    {
      "ply": 0,
      "fen": "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1",
      "maia": {
        "maia_kdd_1100": {
          "policy": { "e2e4": 0.6614, "d2d4": 0.2185 },
          "value": 0.3905
        }
      },
      "stockfish": {
        "depth": 14,
        "cp_vec": { "e2e4": 32, "d2d4": 16 }
      }
    }
  ]
}

Game Type Support

• ✅ Tournament games
• ✅ Lichess games
• ✅ User games (play/hand/brain)
• ❌ Custom PGN/FEN games (excluded - remain local only)

Error Handling

• Network failures logged but don't show user errors
• Invalid data gracefully handled
• Background operations don't block UI
• Proper timer cleanup prevents memory leaks

The implementation is completely transparent to users - analysis is automatically saved and restored without any user intervention.

Fixes #154.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
maia-platform-frontend	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Jul 25, 2025 9:38pm