skills/juliankerignard/mcp-unity/GDD/Gen Agent Trust Hub

GDD

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local project configuration files (unity-planner-status.yaml) and internal skill templates located within the user's local directory (~/.claude/skills/unity-planner/references/templates/gdd-template.md). It saves generated content to a local context directory (.claude/context/unity-planner/). These file operations are strictly local and are essential for the skill's functionality as a documentation tool.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project status data and user-provided gameplay descriptions to generate sections of the GDD. 1. Ingestion points: Reads unity-planner-status.yaml and processes interactive user inputs for game mechanics. 2. Boundary markers: None explicitly defined in the text, although the interactive nature serves as a logical separation. 3. Capability inventory: Writing documentation files to the local project context. 4. Sanitization: No specific sanitization or filtering logic is mentioned for ingested data. This is documented as a processing surface, but no high-risk capabilities are exposed to this data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 04:14 PM
Security Audit — agent-trust-hub — GDD