Description

The test tests/tools/test_vision_tools.py::TestErrorLoggingExcInfo::test_analysis_error_logs_exc_info is failing consistently on the main branch, causing all PR CI checks to fail regardless of the PR's actual changes.

Evidence

• Main branch test runs are failing: https://github.com/NousResearch/hermes-agent/actions?query=branch%3Amain
• Multiple unrelated PRs (fix(context): detect Anthropic 'prompt is too long' as context length error #821, fix(doctor): skip /models health check for MiniMax providers #822, Fix several documentation typos #825, feat: OpenAI-compatible API server — Chat Completions + Responses API #828, fix(gateway): pass model to temporary AIAgent instances #830, Fix/honcho local mode gateway session #832, fix(doctor): add check_url flag to skip /models for providers that don't support it #833) are all failing with the same test
• The failure is:

tests/tools/test_vision_tools.py:257: in test_analysis_error_logs_exc_info
    assert any(r.exc_info is not None for r in error_records)
E   assert False

Impact

This blocks all PR merges until the test is fixed or marked as flaky.

Suggested Fix

Either:

1. Fix the underlying issue in the test/code
2. Mark the test as @pytest.mark.flaky if it's intermittent
3. Skip the test temporarily with @pytest.mark.skip until root cause is identified