common/hermes
2
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

fix(tests): fix 78 CI test failures and remove dead test (#9036)

Browse Source 
Production fixes:
- voice_mode.py: add is_recording property to AudioRecorder (parity with TermuxAudioRecorder)
- cronjob_tools.py: add sms example to deliver description

Test fixes:
- test_real_interrupt_subagent: add missing _execution_thread_id (fixes 19 cascading failures from leaked _build_system_prompt patch)
- test_anthropic_error_handling: add _FakeMessages, override _interruptible_streaming_api_call (6 fixes)
- test_ctx_halving_fix: add missing request_overrides attribute (4 fixes)
- test_context_token_tracking: set _disable_streaming=True for non-streaming test path (4 fixes)
- test_dict_tool_call_args: set _disable_streaming=True (1 fix)
- test_provider_parity: add model='gpt-4o' for AIGateway tests to meet 64K minimum context (4 fixes)
- test_session_race_guard: add user_id to SessionSource (5 fixes)
- test_restart_drain/helpers: add user_id to SessionSource (2 fixes)
- test_telegram_photo_interrupts: add user_id to SessionSource
- test_interrupt: target thread_id for per-thread interrupt system (2 fixes)
- test_zombie_process_cleanup: rewrite with object.__new__ for refactored GatewayRunner.stop() (1 fix)
- test_browser_camofox_state: update config version 15->17 (1 fix)
- test_trajectory_compressor_async: widen lookback window 10->20 for line-shifted AsyncOpenAI (1 fix)
- test_voice_mode: fixed by production is_recording addition (5 fixes)
- test_voice_cli_integration: add _attached_images to CLI stub (2 fixes)
- test_hermes_logging: explicit propagation/level reset for cross-test pollution defense (1 fix)
- test_run_agent: add base_url for OpenRouter detection tests (2 fixes)

Deleted:
- test_inline_think_blocks_reasoning_only_accepted: tested unimplemented inline <think> handling
This commit is contained in:
Teknium
2026-04-13 10:50:24 -07:00
committed by GitHub
parent b909a9efef
commit 0dd26c9495
18 changed files with 92 additions and 50 deletions
Download Patch File Download Diff File Expand all files Collapse all files

21
tests/run_agent/test_anthropic_error_handling.py
View File

@@ -102,7 +102,19 @@ class _PromptTooLongError(Exception):
self.status_code = 400
class _FakeMessages:
"""Stub for client.messages.create() / client.messages.stream()."""
def create(self, **kwargs):
raise NotImplementedError("_FakeAnthropicClient.messages.create should not be called directly in tests")
def stream(self, **kwargs):
raise NotImplementedError("_FakeAnthropicClient.messages.stream should not be called directly in tests")
class _FakeAnthropicClient:
def __init__(self):
self.messages = _FakeMessages()
def close(self):
pass
@@ -131,13 +143,14 @@ def _make_agent_cls(error_cls, recover_after=None):
def run_conversation(self, user_message, conversation_history=None, task_id=None):
calls = {"n": 0}
def _fake_api_call(api_kwargs):
def _fake_api_call(api_kwargs, **kw):
calls["n"] += 1
if recover_after is not None and calls["n"] > recover_after:
return _anthropic_response("Recovered")
raise error_cls()
self._interruptible_api_call = _fake_api_call
self._interruptible_streaming_api_call = _fake_api_call
return super().run_conversation(
user_message, conversation_history=conversation_history, task_id=task_id
)
@@ -352,10 +365,11 @@ def test_401_refresh_fails_is_non_retryable(monkeypatch):
return False # Simulate failed credential refresh
def run_conversation(self, user_message, conversation_history=None, task_id=None):
def _fake_api_call(api_kwargs):
def _fake_api_call(api_kwargs, **kw):
raise _UnauthorizedError()
self._interruptible_api_call = _fake_api_call
self._interruptible_streaming_api_call = _fake_api_call
return super().run_conversation(
user_message, conversation_history=conversation_history, task_id=task_id
)
@@ -436,13 +450,14 @@ def test_prompt_too_long_triggers_compression(monkeypatch):
def run_conversation(self, user_message, conversation_history=None, task_id=None):
calls = {"n": 0}
def _fake_api_call(api_kwargs):
def _fake_api_call(api_kwargs, **kw):
calls["n"] += 1
if calls["n"] == 1:
raise _PromptTooLongError()
return _anthropic_response("Compressed and recovered")
self._interruptible_api_call = _fake_api_call
self._interruptible_streaming_api_call = _fake_api_call
return super().run_conversation(
user_message, conversation_history=conversation_history, task_id=task_id
)

1
tests/run_agent/test_context_token_tracking.py
View File

@@ -56,6 +56,7 @@ def _make_agent(monkeypatch, api_mode, provider, response_fn):
def run_conversation(self, msg, conversation_history=None, task_id=None):
self._interruptible_api_call = lambda kw: response_fn()
self._disable_streaming = True
return super().run_conversation(msg, conversation_history=conversation_history, task_id=task_id)
return _A(model="test-model", api_key="test-key", provider=provider, api_mode=api_mode)

1
tests/run_agent/test_dict_tool_call_args.py
View File

@@ -66,6 +66,7 @@ def test_tool_call_validation_accepts_dict_arguments(monkeypatch):
quiet_mode=True,
skip_memory=True,
)
agent._disable_streaming = True
result = agent.run_conversation("read the file")

15
tests/run_agent/test_provider_parity.py
View File

@@ -44,11 +44,11 @@ class _FakeOpenAI:
pass
def _make_agent(monkeypatch, provider, api_mode="chat_completions", base_url="https://openrouter.ai/api/v1"):
def _make_agent(monkeypatch, provider, api_mode="chat_completions", base_url="https://openrouter.ai/api/v1", model=None):
monkeypatch.setattr("run_agent.get_tool_definitions", lambda **kw: _tool_defs("web_search", "terminal"))
monkeypatch.setattr("run_agent.check_toolset_requirements", lambda: {})
monkeypatch.setattr("run_agent.OpenAI", _FakeOpenAI)
return AIAgent(
kwargs = dict(
api_key="test-key",
base_url=base_url,
provider=provider,
@@ -58,6 +58,9 @@ def _make_agent(monkeypatch, provider, api_mode="chat_completions", base_url="ht
skip_context_files=True,
skip_memory=True,
)
if model:
kwargs["model"] = model
return AIAgent(**kwargs)
# ── _build_api_kwargs tests ─────────────────────────────────────────────────
@@ -247,7 +250,7 @@ class TestBuildApiKwargsChatCompletionsServiceTier:
class TestBuildApiKwargsAIGateway:
def test_uses_chat_completions_format(self, monkeypatch):
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1")
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1", model="gpt-4o")
messages = [{"role": "user", "content": "hi"}]
kwargs = agent._build_api_kwargs(messages)
assert "messages" in kwargs
@@ -255,7 +258,7 @@ class TestBuildApiKwargsAIGateway:
assert kwargs["messages"][-1]["content"] == "hi"
def test_no_responses_api_fields(self, monkeypatch):
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1")
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1", model="gpt-4o")
messages = [{"role": "user", "content": "hi"}]
kwargs = agent._build_api_kwargs(messages)
assert "input" not in kwargs
@@ -263,7 +266,7 @@ class TestBuildApiKwargsAIGateway:
assert "store" not in kwargs
def test_includes_reasoning_in_extra_body(self, monkeypatch):
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1")
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1", model="gpt-4o")
messages = [{"role": "user", "content": "hi"}]
kwargs = agent._build_api_kwargs(messages)
extra = kwargs.get("extra_body", {})
@@ -271,7 +274,7 @@ class TestBuildApiKwargsAIGateway:
assert extra["reasoning"]["enabled"] is True
def test_includes_tools(self, monkeypatch):
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1")
agent = _make_agent(monkeypatch, "ai-gateway", base_url="https://ai-gateway.vercel.sh/v1", model="gpt-4o")
messages = [{"role": "user", "content": "hi"}]
kwargs = agent._build_api_kwargs(messages)
assert "tools" in kwargs

3
tests/run_agent/test_real_interrupt_subagent.py
View File

@@ -76,7 +76,8 @@ class TestRealSubagentInterrupt(unittest.TestCase):
parent._delegate_spinner = None
parent.tool_progress_callback = None
parent.iteration_budget = IterationBudget(max_total=100)
parent._client_kwargs = {"api_key": "test", "base_url": "http://localhost:1"}
parent._client_kwargs = {"api_key": "***", "base_url": "http://localhost:1"}
parent._execution_thread_id = None
from tools.delegate_tool import _run_single_child

23
tests/run_agent/test_run_agent.py
View File

@@ -880,6 +880,7 @@ class TestBuildApiKwargs:
assert kwargs["extra_body"]["reasoning"] == {"enabled": False}
def test_reasoning_not_sent_for_unsupported_openrouter_model(self, agent):
agent.base_url = "https://openrouter.ai/api/v1"
agent.model = "minimax/minimax-m2.5"
messages = [{"role": "user", "content": "hi"}]
kwargs = agent._build_api_kwargs(messages)
@@ -1575,6 +1576,7 @@ class TestHandleMaxIterations:
assert "API down" in result
def test_summary_skips_reasoning_for_unsupported_openrouter_model(self, agent):
agent.base_url = "https://openrouter.ai/api/v1"
agent.model = "minimax/minimax-m2.5"
resp = _mock_response(content="Summary")
agent.client.chat.completions.create.return_value = resp
@@ -1705,27 +1707,6 @@ class TestRunConversation:
assert result["completed"] is True
assert result["api_calls"] == 2
def test_inline_think_blocks_reasoning_only_accepted(self, agent):
"""Inline <think> reasoning-only responses accepted with (empty) content, no retries."""
self._setup_agent(agent)
empty_resp = _mock_response(
content="<think>internal reasoning</think>",
finish_reason="stop",
)
agent.client.chat.completions.create.side_effect = [empty_resp]
with (
patch.object(agent, "_persist_session"),
patch.object(agent, "_save_trajectory"),
patch.object(agent, "_cleanup_task_resources"),
):
result = agent.run_conversation("answer me")
assert result["completed"] is True
assert result["final_response"] == "(empty)"
assert result["api_calls"] == 1 # no retries
# Reasoning should be preserved in the assistant message
assistant_msgs = [m for m in result["messages"] if m.get("role") == "assistant"]
assert any(m.get("reasoning") for m in assistant_msgs)
def test_reasoning_only_local_resumed_no_compression_triggered(self, agent):
"""Reasoning-only responses no longer trigger compression — prefill then accepted."""
self._setup_agent(agent)