Files
Goose-Core/governance/CAPABILITY_GAP.md
mblanke 285de0ce50 Add Capability Map for GooseStrike vs ThreatHunt
This document clarifies the responsibility boundaries between GooseStrike and ThreatHunt, outlining their primary roles, capabilities, control, and risk factors.
2025-12-24 13:06:15 -05:00

47 lines
1.2 KiB
Markdown

# Capability Map — GooseStrike vs ThreatHunt
This document clarifies responsibility boundaries between applications.
## Primary Orientation
| Area | GooseStrike | ThreatHunt |
|----|----|----|
| Primary role | Orchestration & execution | Analysis & investigation |
| User mindset | "What should we do?" | "What does this data mean?" |
| Time focus | Forward-looking | Retrospective / iterative |
---
## Capabilities
| Capability | GooseStrike | ThreatHunt |
|----|----|----|
| Asset discovery | ✅ | ❌ |
| Tool execution | ✅ | ❌ |
| Workflow orchestration | ✅ | ❌ |
| CSV ingestion | ❌ | ✅ |
| Data normalization | ⚠️ (light) | ✅ |
| Deep analysis | ❌ | ✅ |
| Enrichment (VT, intel) | ❌ | ✅ |
| Findings generation | ✅ | ✅ |
| Alerts | ✅ | ⚠️ (derived only) |
---
## Control & Risk
| Area | GooseStrike | ThreatHunt |
|----|----|----|
| Executes actions | Yes | No |
| Requires approvals | Often | No |
| Multi-tenant isolation | Optional | Mandatory |
| Safe for junior analysts | Guarded | Yes |
---
## Rule of Thumb
- **GooseStrike decides and acts**
- **ThreatHunt analyzes and explains**
Overlap is intentional only at the *Finding* layer.