Files
tradingagents/tests/test_anthropic_effort.py
Yijia-Xiao a420ad0f3b fix(cli): honor env precedence for LLM and run config
Interactive selections and flag defaults overrode TRADINGAGENTS_* env vars.
Rule: an explicit env value or CLI flag wins; otherwise the env-applied
default is kept.

- Research depth: skip the prompt when both round-count env vars are set, and
  stop overwriting them (#977).
- Checkpoint: --checkpoint/--no-checkpoint is tri-state; omitting it keeps
  TRADINGAGENTS_CHECKPOINT_ENABLED (#976).
- Docker ollama: use TRADINGAGENTS_LLM_PROVIDER + OLLAMA_BASE_URL, not a bare
  LLM_PROVIDER the overlay never reads (#975).
- Reasoning/thinking knobs: settable via env; the prompt is skipped when set.
- Effort gating: forward effort only to models that accept it (Anthropic
  Opus 4.5+/Sonnet 4.6+, OpenAI reasoning models); drop it elsewhere.
- Boolean env values: raise a named error on invalid input instead of
  silently becoming False.
2026-06-21 21:03:05 +00:00

89 lines
3.3 KiB
Python

"""Tests for Anthropic effort-parameter gating (#831).
Haiku (any version) and Sonnet 4.5 reject the ``effort`` parameter with a
400. Only Opus 4.5+ and Sonnet 4.6+ accept it. The gate uses a per-family
minimum version so future ``claude-{opus,sonnet}-X-Y`` releases inherit
support automatically.
"""
import pytest
from tradingagents.llm_clients import anthropic_client as mod
def _capture_kwargs(monkeypatch):
captured: dict = {}
monkeypatch.setattr(
mod, "NormalizedChatAnthropic",
lambda **kwargs: captured.setdefault("kwargs", kwargs),
)
return captured
@pytest.mark.unit
class TestEffortGate:
@pytest.mark.parametrize(
"model",
[
"claude-haiku-4-5", "claude-haiku-5-0", "claude-haiku-4-7-preview",
# Sonnet 4.5 (and earlier) 400 on effort — only Sonnet 4.6+ supports it.
"claude-sonnet-4-5", "claude-sonnet-4-0",
],
)
def test_unsupported_models_do_not_receive_effort(self, monkeypatch, model):
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(model=model, effort="medium", api_key="x").get_llm()
assert "effort" not in captured["kwargs"]
@pytest.mark.parametrize(
"model",
[
"claude-opus-4-5", "claude-opus-4-6", "claude-opus-4-7",
"claude-sonnet-4-6",
],
)
def test_current_opus_and_sonnet_receive_effort(self, monkeypatch, model):
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(model=model, effort="high", api_key="x").get_llm()
assert captured["kwargs"]["effort"] == "high"
@pytest.mark.parametrize(
"model",
["claude-opus-5-0", "claude-opus-4-8", "claude-sonnet-5-0"],
)
def test_future_opus_sonnet_inherit_effort_via_pattern(self, monkeypatch, model):
"""Forward-compat: new Opus/Sonnet versions don't need a code change."""
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(model=model, effort="low", api_key="x").get_llm()
assert captured["kwargs"]["effort"] == "low"
def test_mythos_preview_receives_effort(self, monkeypatch):
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(
model="claude-mythos-preview", effort="medium", api_key="x"
).get_llm()
assert captured["kwargs"]["effort"] == "medium"
def test_unknown_anthropic_model_does_not_receive_effort(self, monkeypatch):
"""Default is conservative — unknown models don't get effort to avoid 400s."""
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(
model="claude-experimental-x", effort="medium", api_key="x"
).get_llm()
assert "effort" not in captured["kwargs"]
def test_other_kwargs_still_forwarded_when_effort_skipped(self, monkeypatch):
"""Skipping effort must not break other passthrough kwargs."""
captured = _capture_kwargs(monkeypatch)
mod.AnthropicClient(
model="claude-haiku-4-5",
effort="medium",
api_key="placeholder",
max_tokens=1024,
timeout=30,
).get_llm()
assert captured["kwargs"]["api_key"] == "placeholder"
assert captured["kwargs"]["max_tokens"] == 1024
assert captured["kwargs"]["timeout"] == 30
assert "effort" not in captured["kwargs"]