time-tracking

Pass

Audited by Gen Agent Trust Hub on Feb 20, 2026

Risk Level: SAFE
Full Analysis
  • SAFE (SAFE): No malicious patterns or security risks were identified. The skill correctly utilizes 1Password for secret management rather than hardcoding credentials.
  • PROMPT_INJECTION (LOW): Evaluation of indirect prompt injection surface:
  • Ingestion points: User-controlled strings for task descriptions and project names in the scripts/track command.
  • Boundary markers: Instructions use double-quotes in command examples to delimit user-provided text.
  • Capability inventory: Execution of the local script scripts/track, which likely performs authorized API calls to Toggl or Clockify.
  • Sanitization: The skill relies on standard shell quoting in its documentation to prevent basic injection; the internal script implementation is expected to handle string parsing securely.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 20, 2026, 02:59 AM