The Agent Skills Directory

[PROMPT_INJECTION]: The documented workflow involves ingesting external runtime data, which establishes an indirect prompt injection surface.
Ingestion points: insta logs and insta events pull live data from remote cloud services into the agent context as described in cli-reference.md.
Boundary markers: There are no instructions defining boundaries or ignore-directives for data originating from logs or events.
Capability inventory: The agent is provided with instructions for executing impactful commands such as insta deploy and insta project delete as listed in cli-reference.md.
Sanitization: No validation or sanitization processes are mentioned for external log or event data.
[COMMAND_EXECUTION]: The skill requires the agent to execute shell commands via the insta CLI for cloud infrastructure management and includes an auditing tool, insta observe, that installs a hook to monitor agent tool usage.

insta