{
  "context": "Evaluates how well the solution uses pylibmc's compression controls when writing and reading cache entries with per-call overrides for threshold and compression level. Emphasizes correct use of pylibmc write/read APIs with overrides versus defaults.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Set threshold override",
      "description": "`AdaptiveCache.store` calls `pylibmc.Client.set` with `min_compress_len` set from the per-call `min_bytes` argument and falls back to `default_min_bytes` when no override is provided.",
      "max_score": 20
    },
    {
      "name": "Set compression level",
      "description": "`AdaptiveCache.store` passes `compress_level` to `client.set`, using the per-call override when provided and otherwise using the constructor's `default_level` without mutating it between calls.",
      "max_score": 20
    },
    {
      "name": "Bulk overrides",
      "description": "`AdaptiveCache.store_many` uses `client.set_multi` with `min_compress_len` and `compress_level` derived from the provided overrides or defaults and returns the failed keys from `set_multi` unchanged.",
      "max_score": 20
    },
    {
      "name": "Reads via client",
      "description": "`AdaptiveCache.load` retrieves values through `client.get` (or equivalent pylibmc read) and returns the provided default for missing keys without performing manual compression or decompression.",
      "max_score": 20
    },
    {
      "name": "Reuse provided client",
      "description": "All cache operations reuse the injected `pylibmc.Client` instance instead of creating new clients or bypassing pylibmc, ensuring compression settings apply to every write.",
      "max_score": 20
    }
  ]
}