The Agent Skills Directory

[SAFE]: No security issues were detected. The skill follows best practices for tool invocation, such as recommending JSON format for machine parsing and using environment variables for sensitive data like API keys.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes untrusted telemetry data (logs, traces, spans) from external sources.
Ingestion points: Log and trace data retrieved via commands like aspire logs, aspire otel logs, kubectl logs, and az containerapp logs as described in SKILL.md and monitoring.md.
Boundary markers: The instructions do not specify explicit delimiters or warnings to isolate telemetry content from agent instructions.
Capability inventory: The skill utilizes subprocess execution of CLI tools including aspire, kubectl, az, and docker.
Sanitization: No specific sanitization or filtering logic for ingested log content is provided. Note: This risk is inherent to monitoring tools and is considered acceptable for the skill's primary purpose.

aspire-monitoring