The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @googleworkspace/cli package via pnpm. This is a standard procedure for a developer-oriented tool associated with a well-known service.
[COMMAND_EXECUTION]: The skill makes extensive use of the gws CLI to perform operations across Google Drive, Docs, Sheets, and Gmail. It includes safeguards such as a --dry-run flag and instructions for the agent to confirm destructive operations with the user.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads content from external sources (Gmail messages, Google Docs, etc.).
Ingestion points: Data enters the context via gws docs documents get, gws gmail users messages list, and gws drive files export (SKILL.md).
Boundary markers: Not explicitly defined in this file, though external reference files for security rules are mentioned.
Capability inventory: The skill has capabilities to write to the file system, execute CLI commands, and perform network requests via the gws tool.
Sanitization: No explicit sanitization or escaping of the ingested content is detailed in the primary instructions.

google-workspace