detection-scoring
$
npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/detection-scoringRates each failure mode's detectability on an inverted 1-10 scale (10 = hardest to detect).
SKILL.md
.github/skills/detection-scoringView on GitHub ↗
--- name: detection-scoring description: "Rate detectability 1-10 (inverted: 10 = hardest to detect). Estimates how likely current controls would catch the failure before impact." execution: subagent prompt: ./prompt.md input: failure_modes (string), chains (string) used-by: [failure-anticipation] --- # Detection Scoring Rates each failure mode's detectability on an inverted 1-10 scale (10 = hardest to detect). ## Execution Subagent — spawned via subagent-spawning/spawn-agent. ## Why Subagent Detection assessment requires reasoning about observability and monitoring independent of severity/occurrence. Isolated context ensures unbiased evaluation. ## Input - **failure_modes**: Failure mode catalog - **chains**: Effect chains (to assess where detection could occur) ## Output - **scores**: List of (failure_mode_id, detection_score, justification) - **detection_gaps**: Modes with no current detection mechanism