The Agent Skills Directory

[DYNAMIC_EXECUTION]: The generate_script function within scripts/create_datadog_dashboard.py and scripts/create_grafana_dashboard.py allows for the creation of new bash scripts. The content of these scripts is derived from a template parameter which is passed via the command-line argument --content, allowing for arbitrary script generation.
[PRIVILEGE_ESCALATION]: The Python scripts scripts/create_datadog_dashboard.py and scripts/create_grafana_dashboard.py explicitly call file_path.chmod(0o755) on the newly created shell scripts. This automatically grants execution permissions to files generated from potentially untrusted user content.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it translates user-provided dashboard requirements directly into generated files and executable scripts.
Ingestion points: User-supplied requirements for APM dashboards, monitoring dashboards, or performance metrics as described in the SKILL.md triggers.
Boundary markers: Absent. The skill instructions do not include delimiters or directives for the agent to ignore potentially malicious embedded instructions in the user's requirements.
Capability inventory: The skill uses scripts/create_datadog_dashboard.py and scripts/create_grafana_dashboard.py which have file-writing and chmod capabilities.
Sanitization: Absent. The script content template is interpolated directly into the bash script file without escaping or validation.

creating-apm-dashboards