The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads external data from Google Calendar, creating a surface for indirect prompt injection attacks where malicious content in calendar events could influence the agent's behavior.
Ingestion points: The list-events and get-event commands retrieve data such as event summaries, descriptions, and attendee lists from an external API.
Boundary markers: The instructions do not define delimiters or provide warnings to the agent to ignore potentially malicious instructions embedded within the retrieved calendar data.
Capability inventory: The skill possesses write capabilities including create-event, update-event, delete-event, and respond-to-event, which could be abused if an injection is successful.
Sanitization: There is no documentation of content sanitization, filtering, or validation for the data ingested from the Calendar API.
[COMMAND_EXECUTION]: The skill relies on the execution of local Python scripts (scripts/auth.py and scripts/gcal.py) to perform authentication and calendar operations. While these are part of the skill's intended functionality, they represent the primary interface for command execution on the host system.

google-calendar-automation