The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's primary functionality relies on executing a variety of shell commands using a local CLI tool named gbrain. These commands (e.g., gbrain salience, gbrain anomalies, gbrain recall, gbrain waiting) are used to gather data for the daily briefing.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits a vulnerability surface where untrusted data from external sources is interpolated into shell commands.
Ingestion points: Meeting attendee names and email sender names are ingested from the calendar and email systems respectively (defined in SKILL.md).
Boundary markers: There are no boundary markers or instructions to sanitize the external strings before they are used as arguments.
Capability inventory: The skill executes multiple subprocess calls to the gbrain CLI tool with parameters derived from external input.
Sanitization: No sanitization or escaping of the name strings is performed before they are passed to the gbrain search command.

briefing