The Agent Skills Directory

[SAFE]: The skill serves as a documentation and command-routing layer for the gws CLI tool. It defines standard API interactions with Google Drive resources such as files, shared drives, and permissions.
[SAFE]: Although the skill enables reading and writing data to Google Drive, these operations are the primary and intended purpose of the skill. No evidence of unauthorized data exfiltration or hardcoded credentials was found.
[SAFE]: The skill acknowledges security best practices by referring to a shared configuration (../gws-shared/SKILL.md) for authentication and global security rules.
[SAFE]: Indirect Prompt Injection Risk Assessment:
Ingestion points: files.get, files.download, comments.list, replies.list (SKILL.md)
Boundary markers: Not explicitly defined in this file (refers to gws-shared for security rules)
Capability inventory: files.create, files.update, permissions.create, drives.create (SKILL.md)
Sanitization: Not specified in the markdown instructions, assuming agent-level or tool-level handling.

gws-drive