fork-gauge

Pass

Audited by Gen Agent Trust Hub on Apr 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill is purely informational and focused on UI component maintenance.
  • [EXTERNAL_DOWNLOADS]: The component fetches fork risk data from a relative project path (/data/fork-risk.json). This is documented as a standard data flow for the visualization and targets a local asset, posing no risk of untrusted remote code execution.
  • [DATA_EXFILTRATION]: No evidence of unauthorized data transfer or sensitive file access was found. The skill operations are confined to UI rendering and local data ingestion.
  • [PROMPT_INJECTION]: The instructions focus entirely on component maintenance and do not contain patterns designed to subvert agent safety protocols or extract system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 18, 2026, 09:35 AM