mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-30 19:09:28 +00:00
39975613b1
Second, deeper pass over tools/gateway/hermes_cli plus first pass over the trees wave 1 missed (acp, acp_adapter, skills, computer_use, docker, dashboard, conformance, monitoring, secret_sources, hermes_state, providers). Same rubric as wave 1 (AGENTS.md test policy); security, alternation/caching invariants, issue-number regressions, and E2E kept. Real test-quality fixes found and rooted out along the way: - tests/tools/test_command_guards.py made real auxiliary-LLM HTTPS calls (DEFAULT_CONFIG smart-approval leaked in) — pinned approval mode=manual via autouse fixture: 17.4s → 0.4s. - test_model_switch_custom_providers.py / test_user_providers_model_switch.py silently probed live provider catalogs (~2s/test) — stubbed cached_provider_model_ids/provider_model_ids/fetch_api_models. - test_telegram_noise_filter.py: 15-platform copy-paste matrix over shared gateway.run logic → 3 representative platforms (55s → 3.9s). - test_gateway_shutdown.py: stop()'s 5s interrupt-deadline loop spun on MagicMock agents — interrupt.side_effect now clears _running_agents (22s → 1.0s). - test_gateway_inactivity_timeout.py poll-harness timings shrunk 3-5x (24s → 1.1s); test_mcp_stability.py backoff/SIGTERM-grace sleeps patched (15.4s → 2.5s); test_async_delegation.py negative-drain wait 5s → 0.5s. - test_telegram_init_deadline.py: loop-block margin restored to 1.0s with rationale comment — the watchdog-dump assertion needs the loop blocked well past deadline+grace under parallel load (flaked once in the 40-worker verification run at a 0.2s margin). Verification: full hermetic suite via scripts/run_tests.sh — 2,438 files, 21,718 tests passed, 0 failed, 293.9s wall. Suite totals vs original baseline: 46,820 → 19,757 test functions (−57.8%), wall 583.5s → 293.9s (−50%), subprocess CPU 13,564s → 11,623s.
88 lines
3.4 KiB
Python
88 lines
3.4 KiB
Python
"""Tests for #7100 — transient failures (429/timeout) must not drop the
|
|
user message from the transcript.
|
|
|
|
The #1630 fix introduced a blanket skip of transcript writes on any
|
|
``failed`` agent result. That was correct for context-overflow failures
|
|
(which would otherwise cause a session-growth loop), but it also caused
|
|
transient provider failures (rate limits, read timeouts, connection
|
|
resets) to silently drop the user's message — so the agent had no memory
|
|
of the last turn on the next attempt.
|
|
|
|
The gateway classifier must distinguish:
|
|
|
|
* ``compression_exhausted=True`` OR context-keyword errors OR a generic
|
|
``400`` on a long history → context-overflow → skip transcript
|
|
* everything else that fails → transient → persist the user message
|
|
"""
|
|
|
|
|
|
def _classify(agent_result: dict, history_len: int) -> tuple[bool, bool]:
|
|
"""Replicate the gateway classifier from GatewayRunner._run_agent.
|
|
|
|
Returns ``(agent_failed_early, is_context_overflow_failure)``.
|
|
"""
|
|
agent_failed_early = bool(agent_result.get("failed"))
|
|
err = str(agent_result.get("error", "")).lower()
|
|
is_context_overflow_failure = agent_failed_early and (
|
|
bool(agent_result.get("compression_exhausted"))
|
|
or any(p in err for p in (
|
|
"context length", "context size", "context window",
|
|
"maximum context", "token limit", "too many tokens",
|
|
"reduce the length", "exceeds the limit",
|
|
"request entity too large", "prompt is too long",
|
|
"payload too large", "input is too long",
|
|
))
|
|
or ("400" in err and history_len > 50)
|
|
)
|
|
return agent_failed_early, is_context_overflow_failure
|
|
|
|
|
|
class TestContextOverflowStillSkipsTranscript:
|
|
"""#1630 behavior must be preserved for real context-overflow cases."""
|
|
|
|
def test_compression_exhausted_is_context_overflow(self):
|
|
agent_result = {
|
|
"failed": True,
|
|
"compression_exhausted": True,
|
|
"error": "Request payload too large: max compression attempts reached.",
|
|
}
|
|
failed, ctx_overflow = _classify(agent_result, history_len=100)
|
|
assert failed
|
|
assert ctx_overflow
|
|
|
|
|
|
class TestTransientFailureKeepsUserMessage:
|
|
"""Transient provider failures must NOT skip the transcript — doing so
|
|
drops the user message and the agent forgets the turn. (#7100)"""
|
|
|
|
def test_rate_limit_429_is_not_context_overflow(self):
|
|
agent_result = {
|
|
"failed": True,
|
|
"error": (
|
|
"API call failed after 3 retries: 429 Too Many Requests "
|
|
"— rate limit exceeded"
|
|
),
|
|
}
|
|
failed, ctx_overflow = _classify(agent_result, history_len=10)
|
|
assert failed
|
|
assert not ctx_overflow
|
|
|
|
def test_read_timeout_is_not_context_overflow(self):
|
|
agent_result = {
|
|
"failed": True,
|
|
"error": "ReadTimeout: HTTPSConnectionPool(host='api.z.ai'): Read timed out.",
|
|
}
|
|
failed, ctx_overflow = _classify(agent_result, history_len=10)
|
|
assert failed
|
|
assert not ctx_overflow
|
|
|
|
|
|
class TestSuccessfulResultUnaffected:
|
|
def test_successful_result_neither_failed_nor_overflow(self):
|
|
agent_result = {
|
|
"final_response": "Hello!",
|
|
"messages": [{"role": "assistant", "content": "Hello!"}],
|
|
}
|
|
failed, ctx_overflow = _classify(agent_result, history_len=10)
|
|
assert not failed
|
|
assert not ctx_overflow
|