{
  "context": "Evaluates how the solution uses pylibmc to implement single-key caching with per-item TTL and optional compression. Checks that storage and retrieval rely on pylibmc primitives rather than reimplementation, and that compression toggles are correctly wired to pylibmc parameters.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Client init",
      "description": "Constructs one reusable pylibmc.Client with the provided servers (TCP/UDP/UNIX) instead of recreating clients per call, and keeps configuration limited to what TTL/compression needs.",
      "max_score": 20
    },
    {
      "name": "Set with TTL",
      "description": "Stores entries via client.set using the time argument to honor per-call TTL overrides or a default TTL without manual sleep/clock logic.",
      "max_score": 25
    },
    {
      "name": "Get retrieval",
      "description": "Fetches values through client.get (or __getitem__) and uses its default parameter to supply fallback values instead of custom miss handling.",
      "max_score": 15
    },
    {
      "name": "Compression on",
      "description": "When compression is enabled, passes min_compress_len (and optional compress_level) to client.set so large payloads are compressed by pylibmc before storage and transparently decompressed on get.",
      "max_score": 25
    },
    {
      "name": "Compression off",
      "description": "When compression is disabled, sets min_compress_len to 0 (or omits it) so pylibmc stores data uncompressed and still returns it unchanged via get.",
      "max_score": 15
    }
  ]
}