菜单
Use when a user asks to generate, build, or improve a Power BI report/model package from a process request, source description, or business question using the executable report-package path.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when Power BI tenant administration needs Admin API playbooks for workspace inventory, usage metrics, refresh failures, endorsements, sensitivity labels, sharing risk, gateways, and capacity.
Coordinate virtual Power BI expert roles such as BI Architect, DAX Expert, Power Query Engineer, Fabric Architect, Finance BI Lead, QA Reviewer, Security Reviewer, and CoE Lead. Use to replace manual expert routing with role-based analysis and deliverables.
Use when DAX measures require executable or semi-executable unit tests, query-based assertions, XMLA or Tabular Editor validation, edge cases, totals, RLS contexts, and reconciliation evidence.
Use when Power BI work needs PBIP, Git, deployment pipelines, semantic model versioning, Tabular Editor, DAX checks, regression tests, release gates, workspace promotion, or CI/CD governance.
Use when the user asks for Expert Mode, senior Power BI engineering, deep model/report/DAX/M review, precise control, or developer-facing Power BI delivery with explicit technical evidence.
Use when the user asks for Generalisten Mode, Prozess Manager mode, Process Owner mode, autonomous BI delivery, or wants the plugin to handle Power BI work end-to-end without requiring expert knowledge.
| name | powerbi-report-package-runtime |
| description | Use when a user asks to generate, build, or improve a Power BI report/model package from a process request, source description, or business question using the executable report-package path. |
Use this skill as the default executable path when the user asks to generate a Power BI report, build a semantic model, create DAX/Power Query plans, or improve an existing model/report package.
Prefer the repository executor:
python scripts\powerbi_expert_factory.py report-package --process <process-id> --sources "<source description>" --goal "<report goal>" --out outputs\report-packages\<process-id>