Skip to content

Threat diagnosis with OpenAI reasoning published to SOG #85

Description

@CodeBuildder

Context

OpenAI OpenAI model produces structured threat diagnosis for high/critical findings. Parent: #84.

Acceptance Criteria

  • OpenAI model invoked for findings with severity >= high
  • Diagnosis output: threat_type, attack_stage (MITRE ATT&CK phase), recommended_response, confidence
  • Diagnosis written to Sentinel Operations Graph (SOG) as correlated finding referencing original finding_id
  • Reasoning trace stored and queryable from Sentinel Operations Graph (SOG)

Dependencies

Blocked by: none | Blocks: none

Notes

Model tier: OpenAI model.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions