The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external Google Calendar events, which represents a potential attack surface for indirect prompt injection.
Ingestion points: The skill ingests untrusted data through API methods such as events.get, events.list, and calendarList.list as defined in SKILL.md.
Boundary markers: The skill does not define specific delimiters or instructions to the agent to disregard instructions embedded within event fields (e.g., descriptions or titles).
Capability inventory: The skill has extensive write capabilities via the gws binary, including the ability to insert, update, patch, and delete events, calendars, and access control rules (ACLs).
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the calendar API before it is processed by the agent.
[COMMAND_EXECUTION]: The skill relies on the execution of a local command-line interface to interact with Google Workspace.
Evidence: The skill requires the gws binary (as noted in the requires metadata) and instructs the agent to execute commands using the gws calendar and gws schema patterns.

gws-calendar