| name | audit |
| description | Parallel Agent fan-out for deep subsystem research. Use when investigating an unfamiliar subsystem, debugging cross-cutting concerns, or user asks for a deep audit. Takes subsystem name as argument (e.g., /audit reactive-engine, /audit profiler, /audit voice). Manual invocation only — too open-ended for auto-trigger. |
Deep audit of a Logos subsystem using parallel Agent fan-out. Argument: subsystem name.
Known subsystem mappings:
| Argument | Paths to explore |
|---|
voice or hapax-daimonion | agents/hapax_daimonion/, shared/voice*.py |
reactive-engine or engine | logos/engine/, logos/rules/ |
api or logos-api | logos/api/, logos/routes/ |
profiler or profile | agents/profiler/, profiles/, shared/dimensions.py |
axioms or governance | axioms/, shared/axiom_*.py, shared/consent.py |
visual or perception | agents/visual_*/, shared/perception*.py |
frontend or logos-ui | hapax-logos/src/ |
agents | agents/*/, shared/agent_registry.py |
For the given subsystem:
-
Launch 2-3 parallel Agent calls (subagent_type: Explore), each focused on a different facet:
- Agent 1: Code structure, module boundaries, and dependencies
- Agent 2: Test coverage, edge cases, and error handling
- Agent 3: Integration points with other subsystems
-
Synthesize findings into a structured report:
- Architecture overview
- Quality assessment (test coverage, error handling, code smells)
- Integration risks
- Recommended improvements (prioritized)