yogsoth-ai Failure Mode Cataloging
Build a comprehensive failure taxonomy.
Operations
- edge-case-generation → failure-clustering → scaling-regime-detection
Available SOPs
Subagent: edge-case-generation, failure-clustering, scaling-regime-detection Import: web-search
Execution Guidance
Generate edge cases systematically (boundary values, adversarial inputs, distribution shifts, scale extremes), observe which cause failures, cluster failures by mechanism, identify common triggers.
Minimum Yield
<HARD-GATE>
- edge cases generated: >= 20
- failure clusters identified: >= 3
- triggers per cluster: >= 1
</HARD-GATE>
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use |
|---|---|
| deep-insight-web-search | Quick web scanning for landscape understanding. Import of web-browsing/web-search skill. Snippets only — no conclusions from snippets alone. |
| edge-case-generation | Systematically generate boundary inputs — boundary values, adversarial constructions, distribution shifts, rare combinations, scale extremes. |
| failure-clustering | Group observed failures by mechanism (not symptom), identify common triggers per cluster, estimate frequency and severity. |
| scaling-regime-detection | Detect regime changes in scaling behavior — breakpoints where behavior qualitatively shifts, mechanisms behind transitions. |