Skip to content

Latest commit

 

History

History
266 lines (192 loc) · 5.79 KB

File metadata and controls

266 lines (192 loc) · 5.79 KB

Detection Lifecycle

Purpose

The detection lifecycle defines how detection content moves from idea to sustained operational value.

Its purpose is to ensure detections are not created as one-time artifacts, but are instead developed, documented, reviewed, improved, and retired through a structured process.


Lifecycle Objectives

The lifecycle is intended to:

  • create a repeatable path for developing detections
  • improve consistency in content quality
  • support governance and documentation
  • reduce unmanaged detection sprawl
  • clarify when content is ready for broader operational use
  • provide a controlled way to retire outdated or low-value content

Lifecycle Stages

1. Intake

Description

A detection opportunity is identified and recorded.

Common Inputs

  • threat intelligence
  • incident response findings
  • purple team or red team activity
  • threat hunting discoveries
  • ATT&CK coverage gaps
  • SOC analyst feedback
  • known platform or telemetry opportunities
  • mission or compliance requirements

Primary Outputs

  • use case captured
  • problem statement documented
  • initial priority assessed
  • data source needs identified

2. Design

Description

The detection concept is translated into an initial design.

Activities

  • define the behavior to detect
  • identify relevant data sources
  • determine ATT&CK and CKC alignment
  • consider likely false positives
  • estimate expected fidelity
  • define what good triage context should look like

Primary Outputs

  • initial use case logic
  • draft metadata
  • mapping assumptions
  • detection approach documented

3. Build

Description

The detection is created as a repository-managed content artifact.

Activities

  • author rule logic
  • apply metadata schema
  • assign lifecycle state
  • document severity and ownership
  • add supporting notes or template fields

Primary Outputs

  • draft detection content
  • initial repository placement
  • starter documentation

4. Review

Description

The detection is reviewed for correctness, consistency, and completeness.

Review Areas

  • logic quality
  • metadata completeness
  • naming consistency
  • ATT&CK mapping
  • severity selection
  • ownership
  • repository placement
  • triage usefulness

Primary Outputs

  • review comments
  • required corrections
  • improved content quality

5. Validate

Description

The detection is assessed for usefulness and basic accuracy.

Activities

  • confirm query or logic runs successfully
  • review expected behavior coverage
  • assess whether the logic is understandable
  • document assumptions and validation notes
  • confirm the rule aligns with intended use case

Primary Outputs

  • validation notes
  • readiness observations
  • known gaps or open questions

6. Tune

Description

The detection is refined to improve signal quality and reduce unnecessary noise.

Activities

  • document known benign patterns
  • adjust thresholds or exclusions where justified
  • improve triage context
  • refine severity if needed
  • capture known limitations

Primary Outputs

  • updated false-positive notes
  • improved triage guidance
  • clearer operational expectations

7. Promote

Description

The detection advances through lifecycle states based on readiness.

Standard Lifecycle States

  • experimental
  • testing
  • production
  • deprecated

Promotion Considerations

  • metadata completeness
  • documentation quality
  • operational usefulness
  • review completion
  • validation confidence
  • ownership clarity

Primary Outputs

  • lifecycle update
  • improved program visibility
  • tracking matrix updates

8. Operate

Description

The detection is maintained as an active part of the program.

Activities

  • maintain rule quality
  • update supporting documentation
  • capture operational feedback
  • review for continued relevance
  • align with evolving coverage needs

Primary Outputs

  • continued content maintenance
  • updated ownership and notes
  • operational improvement inputs

9. Retire or Replace

Description

The detection is deprecated or replaced when it is no longer appropriate.

Reasons for Retirement

  • obsolete logic
  • duplicate coverage
  • telemetry changes
  • low ongoing value
  • better replacement detection exists
  • content no longer fits program direction

Primary Outputs

  • deprecated lifecycle status
  • documented replacement if applicable
  • preserved historical traceability

Lifecycle States

Experimental

Early-stage content that is drafted and structured but not yet considered mature.

Testing

Content that is under active review and refinement and is being evaluated for broader operational readiness.

Production

Content that is sufficiently documented, governed, and supported for normal operational use.

Deprecated

Content that is no longer recommended for active use and has been retired or replaced.


Lifecycle Principles

The lifecycle should be:

Structured

Each detection should follow a recognizable path.

Visible

Lifecycle state should be clearly documented.

Governed

Movement between states should reflect actual readiness, not just elapsed time.

Maintainable

Detections should continue to be reviewed and improved after creation.

Traceable

Changes, promotions, and retirements should be visible through repository history and program artifacts.


Supporting Repository Artifacts

The lifecycle is supported by:

  • tracking matrix
  • triage guides
  • governance standards
  • change control process
  • exception management process
  • contribution workflow
  • future validation evidence structures

Summary

The detection lifecycle provides the operational path for detection content from initial idea to maintained program asset. It helps ensure detections are built deliberately, documented consistently, and improved over time rather than treated as one-off alerts.