The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection by retrieving and processing untrusted content from Google Chat.\n
Ingestion points: The skill uses gws chat spaces messages and gws chat spaces list (SKILL.md) to ingest message content and space metadata.\n
Boundary markers: Absent. There are no delimiters or specific instructions provided to the agent to treat data from the chat API as untrusted content.\n
Capability inventory: The skill allows for significant actions including deleting spaces (spaces.delete), setting up spaces with members (spaces.setup), and sending messages via the +send helper (SKILL.md).\n
Sanitization: Absent. No filtering, escaping, or validation logic is defined for the content retrieved from external sources before it is interpreted by the agent.\n- [COMMAND_EXECUTION]: The skill facilitates the execution of the gws command-line utility to perform administrative tasks within Google Workspace.\n- [EXTERNAL_DOWNLOADS]: The skill references the official Google Workspace CLI repository for source materials and requires the gws binary to be installed on the host system.

gws-chat