The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources such as Google Calendar event descriptions and Workspace documents. It follows security best practices by using the --wrap-untrusted flag during data retrieval and includes explicit instructions for the agent to treat this content as untrusted and ignore any instructions contained within.
[COMMAND_EXECUTION]: The skill executes the gog command-line utility to fetch workspace data. These executions are constrained to specific subcommands (calendar, drive, docs) and utilize the --readonly flag to prevent unauthorized modifications to the user's account.

gog-meeting-prep