Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) on Model ladder: fix 10 real merged bugs with tests hidden

[
 {
  "duration_api_ms": 33382,
  "stop_reason": "end_turn",
  "session_id": "[REDACTED]",
  "total_cost_usd": 0.010121559999999998,
  "usage": {
   "input_tokens": 24,
   "cache_creation_input_tokens": 20129,
   "cache_read_input_tokens": 302336,
   "output_tokens": 6140,
   "output_tokens_details": {
    "thinking_tokens": 1258
   },
   "server_tool_use": {
    "web_search_requests": 0,
    "web_fetch_requests": 0
   },
   "service_tier": "standard",
   "cache_creation": {
    "ephemeral_1h_input_tokens": 20129,
    "ephemeral_5m_input_tokens": 0
   },
   "inference_geo": "global",
   "iterations": [
    {
     "input_tokens": 2,
     "output_tokens": 570,
     "cache_read_input_tokens": 31872,
     "cache_creation_input_tokens": 796,
     "cache_creation": {
      "ephemeral_5m_input_tokens": 0,
      "ephemeral_1h_input_tokens": 796
     },
     "type": "message"
    }
   ],
   "speed": "standard",
   "fallback_credit": null
  },
  "modelUsage": {
   "claude-haiku-5-5": {
    "inputTokens": 24,
    "outputTokens": 6140,
    "cacheReadInputTokens": 302336,
    "cacheCreationInputTokens": 20129,
    "webSearchRequests": 0,
    "costUSD": 0.010121559999999998,
    "contextWindow": 1000000,
    "maxOutputTokens": 128000,
    "thinkingTokens": 1258,
    "canonicalModel": "claude-haiku-5-5",
    "provider": "firstParty",
    "costBasis": "list"
   }
  },
  "permission_denials": [],
  "terminal_reason": "completed",
  "fast_mode_state": "off",
  "fast_mode_disabled_reason": "sdk_opt_in_required",
  "subagent_stats": {
   "spawned": 0,
   "requested": {
    "background": 0,
    "foreground": 0,
    "unset": 0
   },
   "started_in_background": 0,
   "max_depth": 0,
   "spawned_by_subagents": 0,
   "completed": 0,
   "failed": 0,
   "killed": {
    "parent": 0,
    "user": 0,
    "system": 0
   },
   "refused": {
    "depth_limit": 0,
    "concurrency_limit": 0,
    "budget": 0
   },
   "by_type": {}
  },
  "safety_stops": 0,
  "is_error": false,
  "num_turns": 15,
  "subtype": "success",
  "api_error_status": null,
  "result": "Fixed. A custom wait strategy that evaluates as false is now called on both the sync and async paths, and its return value is slept on. The full suite passes: 185 passed, 1 skipped.\n\n**Root cause:** `_run_wait` in `tenacity/__init__.py` and `AsyncRetrying._run_wait` in `tenacity/asyncio/__init__.py` used `if not self.wait:` to mean \"no wait\". Any falsy strategy, such as an empty list subclass or an object whose `__bool__` returns False, was therefore skipped.\n\n**Change:** Both checks now only treat `None` and numeric zero as \"no wait\":\n\n```python\nif self.wait is None or (isinstance(self.wait, (int, float)) and not self.wait):\n    retry_state.upcoming_sleep = 0.0\n```\n\nNumeric zero still covers the `sum([])` case from the existing comment.\n\n**Tests added:**\n- `tests/test_tenacity.py::test_falsy_wait_strategy_is_called` checks an empty list subclass and an object with `__bool__` returning False, asserting `sleep(1.0)` is called.\n- `tests/test_asyncio.py::test_falsy_wait_strategy_is_called` checks the same for `AsyncRetrying`.\n\nWithout the library fix, the new tests fail for both sync cases and the async case. With the fix, they pass.\n\nI did not run ruff or mypy: the virtualenv has neither installed, so the style and type checks are unverified. The new lines are within the 88-character limit.",
  "ttft_ms": 1938,
  "type": "result",
  "duration_ms": 44247,
  "uuid": "[REDACTED]",
  "ttft_stream_ms": 1400,
  "time_to_request_ms": 260,
  "first_content_frame_ms": 1401,
  "queued_turn_count": 0,
  "result_index": 0
 }
]