The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using the pbi CLI and pipx. It constructs command strings, such as pbi measure create, using DAX expressions and metadata derived from external sources (Excel workbooks and the MEL_DAX_LIBRARY). There is a risk of command injection if these expressions or metadata contain malicious shell metacharacters, as no explicit sanitization or escaping logic is defined.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the pbi-cli-tool package via pipx. This is an external dependency from a non-standard source that is not part of a verified trusted organization or well-known service list.
[DYNAMIC_EXECUTION]: In the 'Full authoring' route, the skill generates and suggests the execution of a Python script (gen_cerv_dashboard_pbip.py) to programmatically build complex directory structures and files. This runtime generation of executable code increases the attack surface.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from Excel workbooks (sheet names, headers, and values) and local Power BI project files. This data is used to drive automated logic and command construction.
Ingestion points: Excel files read via openpyxl and .pbip metadata files.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the prompt generation logic.
Capability inventory: File system write access and execution of the pbi CLI tool.
Sanitization: No evidence of input validation or character escaping for data interpolated into shell commands.

power-bi-build