{
  "context": "Evaluates whether the solution uses CPython's standard HTTP and URL tooling to implement the snapshot fetcher. It checks that query merging, headers, HEAD guarding, redirect limits, and timeout handling are wired through the stdlib primitives rather than ad-hoc code. The focus is entirely on correct use of urllib and http.client behaviors.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "Query merge",
      "description": "Builds request URLs by parsing existing queries with urllib.parse.urlparse/parse_qsl and re-encoding merged pairs via urllib.parse.urlencode + urlunparse before issuing the HTTP call.",
      "max_score": 20
    },
    {
      "name": "Headers",
      "description": "Constructs requests with urllib.request.Request (or http.client connection headers) to set a default User-Agent and cleanly override it when custom headers are provided.",
      "max_score": 20
    },
    {
      "name": "HEAD guard",
      "description": "Performs a preliminary HEAD using urllib.request.Request(method=\"HEAD\") or http.client.HTTPConnection.request(\"HEAD\", ...) and compares Content-Length to max_bytes, aborting before GET when over budget.",
      "max_score": 20
    },
    {
      "name": "Redirect limit",
      "description": "Follows redirects with urllib.request.build_opener/HTTPRedirectHandler or manual Location header loops, capping at the specified redirect_limit and surfacing an error when the cap is exceeded.",
      "max_score": 20
    },
    {
      "name": "Timeouts",
      "description": "Passes the timeout argument through urllib.request.urlopen or http.client.HTTP(S)Connection to enforce the specified wait window and propagates socket.timeout/URLError when servers stall.",
      "max_score": 20
    }
  ]
}