yogsoth-ai Mitigation Validation Tactic
Validate that proposed mitigations do not introduce new failure modes — a mini-FMEA on the mitigations themselves.
Orchestration
- Receive proposed mitigation measures from mitigation-design-sop
- failure-mode-extraction identifies potential failure modes of each mitigation:
- Could the prevention measure fail?
- Could the detection mechanism produce false negatives?
- Could the response plan create new problems?
- severity-scoring rates new failure modes
- If any new mode scores H-priority:
- Flag mitigation as risky
- mitigation-design-sop redesigns or adds safeguards
- Re-validate (max 2 iterations to prevent infinite loops)
- re-scoring confirms final S/O/D with validated mitigations
Iteration Control
- Max validation iterations: 2 (prevent infinite recursion)
- If still H-priority after 2 iterations: escalate to human review
- Document residual risk for accepted mitigations
Subagents Dispatched
- failure-mode-extraction (mitigation failure identification)
- severity-scoring (new risk assessment)
- mitigation-design-sop (redesign if needed)
- re-scoring (final confirmation)
Termination Conditions
- All mitigations validated as not introducing H-priority risks
- Max iterations reached (escalate with documentation)
- Residual risk accepted and documented
Available SOPs
Optional, no fixed order; the final leaf is always a sop.
| SOP | When to use |
|---|---|
| failure-mode-extraction | Extract structured failure mode list from raw scenarios or artifact analysis. Produces standardized failure mode records. |
| mitigation-design-sop | Design prevention, detection, and response measures for high-priority failure modes. Produces actionable countermeasure specifications. |
| re-scoring | Re-evaluate S/O/D scores after mitigation measures are in place. Validates that mitigations actually reduce risk as expected. |
| severity-scoring | Rate failure mode severity 1-10 based on end-effect impact. Follows AIAG-VDA severity scale calibrated for research artifacts. |