Pending: Skill-based (/smoke-test) or script-based (pytest)?
Pending: Run automatically or on-demand?
User Tasks
FR-051: System Integration Testing
Summary
A /smoke-test skill and test suite that verifies skills, hooks, agents, rules, and templates work correctly together.
Problem / Motivation
FR-049 covers Python unit testing. But the Opus system is more than Python code — it’s a constellation of skills, hooks, agents, rules, templates, and vault conventions that must work together. Currently there’s no way to verify that:
Skills produce valid output and don’t error
Hooks fire correctly and don’t block valid operations
Templates contain all required sections
Rules load for the right paths
Frontmatter schemas are consistent across all FR files
Cross-references between files are valid (no broken links)
As the system grows, changes to one component can silently break others.
Proposed Solution
Two layers:
/smoke-test skill — quick health check runnable in any session
Automated test suite — deeper validation, runnable via script
Test Categories
Category
What it checks
Template integrity
All templates have required sections, valid YAML frontmatter
FR consistency
All FRs have valid frontmatter, status matches directory, no orphan references
Hook health
Hook scripts exist, are executable, handle valid input without error