The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches and processes runtime logs from external cloud environments using the mcloud logs tool, which serves as a potential vector for indirect instructions if log content is untrusted.
Ingestion points: The output of the mcloud logs command is the primary entry point for external data into the agent's context.
Boundary markers: The skill documentation does not define specific delimiters or instructions for the agent to ignore potentially adversarial content within the logs.
Capability inventory: The skill uses the Bash tool to execute mcloud and jq commands.
Sanitization: There are no specific sanitization or filtering steps mentioned to clean the log data before it is read by the agent.

mcloud-logs