The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data and metadata from active Power BI models, which represents an attack surface for indirect prompt injection if those models contain untrusted content designed to influence the agent's behavior.
Ingestion points: The workflow defined in SKILL.md uses model_operations, table_operations, relationship_operations, and measure_operations to connect to and extract the state of current semantic models.
Boundary markers: The instructions lack specific delimiters or

powerbi-modeling