Problem
The monitoring system — the core "readjustment" pillar of OmoiOS — has zero integration tests:
MonitoringLoop — Orchestrates Guardian + Conductor loops
IntelligentGuardian — LLM-powered per-agent trajectory analysis
ConductorService — System-wide coherence scoring + duplicate detection
These are critical services that determine if agents are drifting from goals.
What to do
Create test files in backend/tests/integration/:
test_monitoring_loop.py: Test the loop starts, runs Guardian/Conductor cycles, and handles errors
test_intelligent_guardian.py: Test trajectory analysis scoring, intervention decisions, mock LLM responses
test_conductor.py: Test coherence scoring, duplicate work detection
Use pytest-asyncio for async tests and mock the LLM service to avoid real API calls.
Reference: ARCHITECTURE.md — Testing Requirements checklist
Difficulty: Medium-Hard | Skill: Python (pytest, async testing)
Problem
The monitoring system — the core "readjustment" pillar of OmoiOS — has zero integration tests:
MonitoringLoop— Orchestrates Guardian + Conductor loopsIntelligentGuardian— LLM-powered per-agent trajectory analysisConductorService— System-wide coherence scoring + duplicate detectionThese are critical services that determine if agents are drifting from goals.
What to do
Create test files in
backend/tests/integration/:test_monitoring_loop.py: Test the loop starts, runs Guardian/Conductor cycles, and handles errorstest_intelligent_guardian.py: Test trajectory analysis scoring, intervention decisions, mock LLM responsestest_conductor.py: Test coherence scoring, duplicate work detectionUse
pytest-asynciofor async tests and mock the LLM service to avoid real API calls.Reference:
ARCHITECTURE.md— Testing Requirements checklistDifficulty: Medium-Hard | Skill: Python (pytest, async testing)