The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains instructional examples of powerful system-level commands used for fault injection, such as tc for network latency, iptables for traffic blocking, kubectl for pod termination, and stress-ng for resource saturation. These are provided as code blocks for educational purposes within the context of resilience testing.
[PROMPT_INJECTION]: The skill includes instructions to ingest information from an external file, .agents/qa-project-context.md. While this constitutes a surface for indirect prompt injection, the skill uses this data for project-specific context and does not include instructions that attempt to bypass safety filters or override agent core behavior.
[SAFE]: The skill provides comprehensive safety documentation, including anti-patterns, mandatory steady-state hypotheses, and explicit requirements for monitoring and rollback procedures, which mitigates the risk of the included commands being used destructively.

chaos-engineering