The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data such as container logs and inspection output which may contain malicious instructions designed to influence the agent. * Ingestion points: Reads output from docker logs, docker compose logs, and docker inspect in SKILL.md and scripts/sk-container-ps.sh. * Boundary markers: The Quality Standards and Evidence Produced sections in SKILL.md explicitly instruct the agent to provide redacted logs. * Capability inventory: The skill has the capability to execute administrative commands including docker run, docker exec, and systemctl as described in the workflow section of SKILL.md. * Sanitization: No automated sanitization is implemented in the provided shell script for log content.

linux-container-deployment