The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill creates a surface for indirect prompt injection by ingesting untrusted data from external sources.
Ingestion points: Untrusted data enters the agent context via the gog sheets get command (found in SKILL.md).
Boundary markers: None identified; instructions do not specify delimiters to separate spreadsheet data from system instructions.
Capability inventory: The skill has significant write capabilities including gog sheets update, gog sheets create, and gog drive move (SKILL.md).
Sanitization: There is no evidence of sanitization or validation of the content retrieved from the spreadsheet before it is processed by the agent.
[External Downloads] (LOW): The skill depends on a non-standard CLI tool gog (v0.10.0). While installation is not handled within the skill, the tool is not part of the trusted organization/repository list and its provenance cannot be verified from the provided content.
[Command Execution] (LOW): The skill frequently uses the exec tool to run shell commands with variable interpolation (e.g., $ID, $RESULT). While necessary for the skill's purpose, this pattern requires the agent to properly sanitize inputs to prevent command injection if variables are derived from untrusted spreadsheet data.

google-sheets-ops