tsconfig-inheritance-flattener-mcp
Resolves TypeScript config inheritance chains to reveal effective compiler options, resolves path aliases to physical files, and validates project references for AI agents.
README
🔍 tsconfig-inheritance-flattener-mcp
Your AI agent reads tsconfig.json. It has no idea what it actually means.
MCP server that resolves the full TypeScript config inheritance chain and returns the effective compiler options that actually apply — including everything inherited from extended base configs, monorepo packages, and node_modules presets.
🤔 The problem
Your agent reads tsconfig.json and sees:
{ "extends": "@tsconfig/strictest", "compilerOptions": { "paths": { "@/*": ["./src/*"] } } }
It has no idea that @tsconfig/strictest sets strict: true, noUncheckedIndexedAccess: true, exactOptionalPropertyTypes: true. It doesn't know that baseUrl is defined two levels up in your monorepo base config. So it:
- Suggests code that would fail
noUncheckedIndexedAccess - Gets confused about what
@/resolves to - Doesn't know your
targetisES2022, notES5 - Gives wrong answers about module resolution
The TypeScript compiler API already resolves all of this. This MCP just exposes it.
🛠️ Tools
get_effective_compiler_options
Resolves the full extends chain and returns the merged compiler options that actually apply to a given tsconfig.json. Shows the inheritance chain, all merged options (with enums as readable strings, not magic numbers), and include/exclude patterns.
Effective TypeScript Configuration
Config: /project/apps/web/tsconfig.json
Inheritance chain: /project/apps/web/tsconfig.json
→ /project/tsconfig.base.json
→ node_modules/@tsconfig/strictest/tsconfig.json
Compiler Options (merged):
target: "ES2022"
module: "NodeNext"
moduleResolution: "NodeNext"
strict: true
noUncheckedIndexedAccess: true
exactOptionalPropertyTypes: true
baseUrl: "/project"
paths: { "@/*": ["apps/web/src/*"] }
resolve_module_alias
Maps a TypeScript path alias (e.g. @/hooks/useAuth) to its physical file location on disk, using the resolved paths and baseUrl from the tsconfig. Returns all existing candidates with extension probing.
Alias Resolution: @/hooks/useAuth
Config: /project/apps/web/tsconfig.json
Base URL: /project
Resolved physical paths:
/project/apps/web/src/hooks/useAuth.ts ✓ exists
analyze_project_references
Inspects the references array in a root tsconfig.json and validates that each referenced package has composite: true. Catches broken cross-package dependencies in TypeScript monorepos before they cause silent build failures.
Project References Analysis
Config: /project/tsconfig.json
References found: 2
[✓] packages/shared → /project/packages/shared/tsconfig.json
[✗ NOT FOUND] packages/deprecated → /project/packages/deprecated/tsconfig.json
Violations:
✗ packages/shared is referenced but does not have composite: true
Fix: add "composite": true to packages/shared/tsconfig.json
🧪 What it looks like in practice
Agent is helping debug a TypeScript error and asks:
"What compiler options are actually active in this project?"
Without this MCP, the agent guesses based on what it sees in tsconfig.json. With it:
get_effective_compiler_options("/project/apps/web/tsconfig.json")
→ strict: true, noUncheckedIndexedAccess: true, target: "ES2022", module: "NodeNext"
Now the agent knows exactly why arr[0] has type string | undefined and not just string. No more wrong suggestions.
⚡ Setup
{
"mcpServers": {
"tsconfig-flattener": {
"command": "npx",
"args": ["-y", "tsconfig-inheritance-flattener-mcp"]
}
}
}
🚀 Usage
"What compiler options actually apply to
/project/apps/web/tsconfig.json? It extends a monorepo base and @tsconfig/strictest."
"Where does
@/components/Buttonresolve to on disk?"
"Are the project references in my root tsconfig valid? Do all referenced packages have composite: true?"
Works great alongside:
- ast-impact-mapper-mcp — for code→test correlation
- release-readiness-triage-mcp — for CI triage
📦 Links
- npm: npmjs.com/package/tsconfig-inheritance-flattener-mcp
- GitHub: github.com/vola-trebla/tsconfig-inheritance-flattener-mcp
License
MIT
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。