macos-calendar

Pass

Audited by Gen Agent Trust Hub on Aug 19, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied text for calendar event summaries and descriptions.
  • Ingestion points: User input is parsed into a JSON object which is then processed by scripts/calendar.sh.
  • Boundary markers: The skill requires structured JSON input and uses a Python validation layer to enforce data types.
  • Capability inventory: The skill can create, list, and modify events in the macOS Calendar app and write to a local log file.
  • Sanitization: String inputs are safely handled by passing them as arguments to osascript via the argv array rather than string interpolation, which effectively prevents AppleScript injection attacks.
  • [COMMAND_EXECUTION]: The script executes local system binaries including osascript, python3, pgrep, and open. These are used for the intended purpose of interacting with the macOS Calendar application and parsing JSON data.
  • [DATA_EXPOSURE]: The skill maintains a local activity log at logs/calendar.log. It does not perform any network operations or exfiltrate data to external servers.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 19, 2026, 07:45 PM
Security Audit — agent-trust-hub — macos-calendar