The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates and executes shell commands using gcloud and jq. Environment and user-supplied variables such as $GCP_PROJECT, $TOPIC_NAME, $SUB_NAME, and $SCHEMA_NAME are interpolated into these commands to manage Pub/Sub resources.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data retrieved from the Google Cloud environment, such as resource names, labels, and schema definitions.
Ingestion points: Data enters the agent's context through the output of discovery commands like gcloud pubsub topics list and gcloud pubsub schemas describe in SKILL.md.
Boundary markers: None specific; the skill relies on standard shell execution and agent interpretation.
Capability inventory: The skill possesses the capability to execute shell commands (bash) and interact with GCP services via gcloud CLI.
Sanitization: The skill uses jq to parse JSON output from CLI commands, which provides structural validation of the external data before it is presented to the agent.

managing-google-pubsub