name: cyber-detection-engineering
guardrail: true
description: Designs and reviews detections from telemetry prerequisites, correlation logic, Sigma metadata, severity, status, and benign cases. Use for detection rules, Sigma review, analytic coverage, or tuning; not generic logging advice or live deployment.
metadata:
triggers:
files: []
keywords:
- detection engineering
- detection rule
- Sigma rule
- analytic coverage
- telemetry prerequisite
- false positive tuning
Cyber Detection Engineering
Priority: P0 (CRITICAL)
Do not claim analytic coverage without required telemetry and executable support.
Structure
Input: analytic question, behavior hypothesis, data source/fields, retention, clock quality, scope. Process: define predicates, entity keys, window, cardinality, ordered correlation, suppression, benign cases. Output: rule/spec, prerequisites, severity, lifecycle status, fixtures, limitations, evidence record.
Rules
- A missing source, field, retention, clock, or collection path makes coverage
blocked or not-tested.
- Multi-step behavior requires ordered sequence correlation over related entities/events; never substitute Boolean OR across unrelated single events.
- Sigma
level expresses impact/priority (informational, low, medium, high, critical). Sigma status expresses lifecycle (stable, test, experimental, deprecated). A valid portable rule does not prove backend execution or telemetry.
- Record expected matches, exclusions and benign/false-positive cases with rationale.
- Include scope reference; skill/version/source; observation time; status; evidence references; limitations; owner. Reuse shared skill IDs.
- Production queries or modifications require documented authorization and runtime-proven controls; offline fixtures only.
Anti-Patterns
- Never write
severity: high as Sigma metadata; use level: high.
- Never equate
level with status, rule validity with execution, or synthetic matches with efficacy.
- Never hide missing telemetry by lowering priority or widening OR logic.
References (lazy, primary)