The Agent Skills Directory

[SAFE]: No evidence of prompt injection, obfuscation, credential harvesting, or unauthorized network operations was found in the provided documentation.
[NO_CODE]: The skill description mentions three Python scripts—slo_designer.py, alert_optimizer.py, and dashboard_generator.py—that are intended to perform the core logic. However, these files were not included in the source provided for analysis, preventing a security audit of the executable logic.
[SAFE]: The skill is designed to process user-supplied configuration data (JSON files for services and alerts). This represents a standard functional data ingestion surface. Because no executable capabilities or unsafe interpolation patterns were identified in the instructions, this surface does not present a measurable risk of indirect prompt injection.

observability-designer