Coding-Hard-Bloom-Filter

The model receives the prompt (and optional system message). The run uses scorer javascript_function_tests with the JSON configuration below. Pass/fail and partial credit are determined entirely by that scorer against the model output; no human grading.

Return JSON only with a string field named code. The code must be dependency-free JavaScript, define the requested function in the top level or module.exports, and include no markdown, imports, require, timers, network, filesystem, eval, or placeholders.

Implement function createBloomFilter(capacity, script). Create a probabilistic bloom filter backed by a bit array of size capacity. Use exactly 3 internal hash functions derived from the item string (e.g. simple polynomial hashes with different seeds modulo capacity). The script is a comma-separated string of operations: 'add:VALUE' inserts VALUE into the filter (returns null), 'has:VALUE' tests membership (returns true or false). Return an array of results in operation order (null for add, boolean for has).

Your JSON must look like {"code":"function createBloomFilter(...) { ... }"}.

{
  "function_name": "createBloomFilter",
  "timeout_ms": 100,
  "test_pass_threshold": 1,
  "partial_credit_threshold": 0.5,
  "tests": "[hidden executable tests]"
}