ralph-specum-cancel

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security risks detected in the skill logic.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads from .ralph-state.json and .current-spec. These are localized metadata files for the Ralph Specum system and do not contain or exfiltrate sensitive user credentials or system data.
  • [COMMAND_EXECUTION]: The skill performs directory and file deletions as part of its 'cancel' and 'remove' functionality. These operations are restricted to project-specific paths and require user confirmation for directory removal.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 07:36 PM
Security Audit — agent-trust-hub — ralph-specum-cancel