The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill interfaces with external data sources that may contain malicious instructions.
Ingestion points: Data enters via helper commands that read Gmail messages, Google Calendar events, and Google Drive files.
Boundary markers: The skill lacks explicit markers or instructions to isolate untrusted content from the system prompt.
Capability inventory: The skill uses the gws command-line tool which has capabilities to read from and write to various Google Workspace services.
Sanitization: The provided instruction file does not include sanitization or validation logic for data retrieved from external sources.

gws-workflow