The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous example commands for standard industry tools such as kubectl, aws-cli, and curl. These are used appropriately within the context of monitoring, incident response, and tool configuration.
[CREDENTIALS_UNSAFE]: The skill demonstrates safe secret management by using environment variable placeholders (e.g., ${DD_API_KEY}) and explicitly advising users to replace them with actual keys or use environment variables in its templates and scripts.
[DYNAMIC_EXECUTION]: The dashboard_generator.py script dynamically creates Grafana dashboard configurations in JSON format. This process is handled using standard JSON serialization and does not involve risky code execution like eval() or exec().
[INDIRECT_PROMPT_INJECTION]: The log_analyzer.py script parses external log files. However, its functionality is limited to regex matching, frequency analysis, and pattern extraction. It does not interpolate log content into AI prompts or use it to construct shell commands, effectively mitigating injection risks.
[EXTERNAL_DOWNLOADS]: The documentation references official installation scripts from well-known services (e.g., Datadog's install script) and standard package managers (Helm, Pip), which is expected for an observability implementation skill.

monitoring-observability