PBIP Builder MCP Server
Creates, inspects, validates, and modifies Power BI Project (.pbip) folders, generating PBIR-style reports and TMDL semantic models from structured inputs.
README
PBIP Builder MCP Server
PBIP Builder MCP Server is a local TypeScript MCP server that helps AI coding agents create, inspect, validate, and modify Power BI Project (.pbip) folders. It generates PBIR-style report folders, TMDL semantic model files, report pages, and templated visuals from structured inputs instead of asking an agent to free-write report JSON.
What It Builds
Generated projects follow the Power BI Desktop developer mode layout:
SalesDashboard/
├── SalesDashboard.pbip
├── SalesDashboard.Report/
│ ├── .platform
│ ├── definition.pbir
│ └── definition/
│ ├── version.json
│ ├── report.json
│ └── pages/
│ ├── pages.json
│ └── <PageId>/
│ ├── page.json
│ └── visuals/<VisualId>/
│ ├── visual.json
│ └── mobile.json
└── SalesDashboard.SemanticModel/
├── .platform
├── definition.pbism
└── definition/
├── database.tmdl
├── model.tmdl
├── relationships.tmdl
├── roles/
└── tables/<TableName>.tmdl
Install
npm install
npm run build
Run The MCP Server
npm run dev
The server uses stdio transport. A VS Code MCP configuration is included in .vscode/mcp.json.
To restrict writes, set PBIP_ALLOWED_WORKSPACES to one or more workspace roots separated by the OS path delimiter. If unset, writes are limited to process.cwd().
Microsoft JSON Schemas
The schema registry looks for Microsoft schemas under schemas/microsoft-json-schemas. Add a local copy of the Microsoft json-schemas repository there:
git clone https://github.com/microsoft/json-schemas schemas/microsoft-json-schemas
When schemas are unavailable, validation still checks project structure, JSON parsing, page/visual references, and semantic field bindings, and returns schema warnings rather than failing the project.
Tools
Project tools: pbip_create_project, pbip_open_project, pbip_describe_project, pbip_validate_project, pbip_delete_project, pbip_export_summary.
Semantic model tools: model_add_table, model_update_table, model_delete_table, model_add_column, model_add_measure, model_update_measure, model_delete_measure, model_add_relationship, model_delete_relationship, model_add_role, model_generate_from_schema, model_describe, model_validate.
Report/page/visual tools: report_create, report_update_settings, report_set_theme, report_validate, page_add, page_update, page_delete, page_rename, page_set_size, page_list, visual_add, visual_update, visual_delete, visual_set_position, visual_bind_fields, visual_set_format, visual_set_filter, visual_list, visual_describe.
Blueprint and schema tools: blueprint_generate_project, blueprint_apply, blueprint_validate, blueprint_preview, schema_list, schema_get, schema_validate_json, schema_sync, schema_explain_error.
Example
npm run generate:example
This creates examples/generated/SalesDashboard from examples/sales-dashboard-blueprint.json, including one semantic model table, one report page, and five visual types: card, clustered column chart, table, slicer, and line chart.
Testing
npm test
Tests cover project creation, dry runs, path safety, semantic model generation, page generation, visual generation, blueprint generation, and validation.
Limitations
This is the MVP local generator. Fabric publishing, refresh, live data source setup, custom visuals, advanced formatting, page duplication, visual duplication, and automatic Power BI Desktop repair workflows are intentionally left for later phases. Final compatibility should still be confirmed by opening generated .pbip files in Power BI Desktop.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。