Initial commit
This commit is contained in:
40
skills/using-system-archaeologist/.archive/README.md
Normal file
40
skills/using-system-archaeologist/.archive/README.md
Normal file
@@ -0,0 +1,40 @@
|
||||
# Archive of Untested Briefings
|
||||
|
||||
This directory contains briefings that were created without following the proper RED-GREEN-REFACTOR TDD cycle.
|
||||
|
||||
## Untested Briefings (v1)
|
||||
|
||||
Created: 2025-11-19
|
||||
Created by: Maintenance workflow without behavioral testing
|
||||
Violation: Writing-skills Iron Law - "NO SKILL WITHOUT A FAILING TEST FIRST"
|
||||
|
||||
### Files
|
||||
|
||||
1. **assessing-code-quality-v1-untested.md** (~400 lines)
|
||||
- Content coverage: Complexity, duplication, code smells, maintainability, dependencies
|
||||
- Problem: No baseline testing to verify agents follow guidance
|
||||
- Use: Reference for content areas to cover in tested version
|
||||
|
||||
2. **creating-architect-handover-v1-untested.md** (~400 lines)
|
||||
- Content coverage: Handover report generation, consultation patterns, architect integration
|
||||
- Problem: No baseline testing to verify agents follow guidance
|
||||
- Use: Reference for content areas to cover in tested version
|
||||
|
||||
## Purpose
|
||||
|
||||
These files are archived (not deleted) to:
|
||||
- Track what content areas should be covered
|
||||
- Compare tested vs. untested versions
|
||||
- Document the improvement from proper TDD methodology
|
||||
- Serve as reference when designing pressure scenarios
|
||||
|
||||
## Tested Versions
|
||||
|
||||
Properly tested versions (RED-GREEN-REFACTOR) will be created in the parent directory following:
|
||||
1. **RED:** Baseline scenarios WITHOUT skill - document exact failures
|
||||
2. **GREEN:** Write minimal skill addressing observed rationalizations
|
||||
3. **REFACTOR:** Find loopholes, plug them, re-test until bulletproof
|
||||
|
||||
## Do Not Use
|
||||
|
||||
These untested files should NOT be used in production. They have not been validated through behavioral testing with subagents and may contain gaps, rationalizations, or ineffective guidance.
|
||||
Reference in New Issue
Block a user