apple-calendar
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from calendar events which creates a potential surface for indirect prompt injection.
-
- Ingestion points: Event summaries, descriptions, locations, and URLs are read into the agent context in
scripts/cal-read.sh,scripts/cal-events.sh, andscripts/cal-search.sh.
- Ingestion points: Event summaries, descriptions, locations, and URLs are read into the agent context in
-
- Boundary markers: The skill output uses pipe delimiters (|) but lacks explicit instructions for the agent to ignore embedded commands within the event fields.
-
- Capability inventory: The skill can create, update, and delete calendar events via
osascript.
- Capability inventory: The skill can create, update, and delete calendar events via
-
- Sanitization: No sanitization or filtering is performed on the data retrieved from Calendar.app before it is presented to the agent.
- [COMMAND_EXECUTION]: The skill uses
osascriptto interact with the macOS Calendar application. The bash scripts pass arguments to the AppleScriptrunhandler using the `osascript - args` pattern, which is a safe practice that prevents data from being interpreted as code.
Audit Metadata