The Agent Skills Directory

[SAFE]: The skill is a collection of technical guidelines and reference files for generating queries in the OPAL domain-specific language. Analysis across all 11 threat categories found no evidence of malicious patterns, obfuscation, or unauthorized data access attempts.
[COMMAND_EXECUTION]: The skill instructs the agent to use the observe CLI tool for loading reference documentation. As the skill is authored by 'observeinc', this is considered legitimate vendor-specific functionality for the Observe platform.
[INDIRECT_PROMPT_INJECTION]: The skill translates user requests and dataset metadata into queries, which is an inherent attack surface. This is addressed by instructions that enforce strict schema adherence and type-safe data handling practices. (Ingestion points: User-supplied natural language queries and dataset schemas from the platform; Boundary markers: Structured multi-step workflow and mandatory reference loading; Capability inventory: Generation of OPAL pipelines and execution of the observe CLI tool; Sanitization: Explicit rules for type casting (e.g., string(), float64()) and validation against known schema fields).

generate-opal