SharpLens

io.github.pzalutski-pixel由 GitHub 账号 pzalutski-pixel 发布v1.5.3更新于 Oct 26, 2026

67 semantic C#/.NET analysis tools via Roslyn: navigation, refactoring, codegen, audit.

已验证STDIO仅桌面开发者工具数据与分析

概览

AI 生成的概览

本地 MCP 服务器,借助 Roslyn 为 AI 助手提供 C#/.NET 解决方案的语义分析:导航、重构、代码生成与代码审计。

可以做什么
  • 导航与搜索符号:定义、引用、实现、调用方、调用图、类型层次、重载、特性与 XML 文档
  • 分析代码:编译器与分析器诊断、变更影响、数据流与控制流、复杂度指标、死代码、循环依赖、公共 API 差异
  • 重构与生成:重命名、修改签名、提取方法或接口、整理 using、格式化、应用代码操作、添加空值检查、相等性成员与测试桩
  • 质量审计:上帝对象、未被测试的代码、异步问题、架构规则、命名违规、未使用的依赖
  • 检查工作区:加载解决方案、同步已编辑文件、查看项目结构、依赖图、NuGet 引用与生成的代码
适用场景
适合助手在 C#/.NET 代码库上工作、需要纯文本搜索或 LSP 无法给出的编译器级准确答案时使用,例如影响分析、死代码、复杂度指标,或带预览的解决方案级安全重构。
运行要求
以 stdio 在本地运行,可通过 .NET 全局工具或 npx 安装。需要 .NET 8.0 SDK 或更高版本以及支持 MCP 的客户端。需将 DOTNET_SOLUTION_PATH 设为 .sln、.slnx 或 .slnf 文件,或先调用 load_solution。可选设置包括 SHARPLENS_LOG_LEVEL、SHARPLENS_TIMEOUT_SECONDS、SHARPLENS_MAX_DIAGNOSTICS、SHARPLENS_ABSOLUTE_PATHS 与 SHARPLENS_ENABLE_SEMANTIC_CACHE。未声明任何身份验证。
安装前请注意
许多工具会写入源文件:重构、代码修复、格式化与生成类工具会修改 .cs 文件,建议使用预览模式并配合版本控制。在服务器之外改动的文件必须用 sync_documents 同步,否则结果会过期。服务器在内存中保存解决方案副本,加载解决方案会执行 MSBuild 还原,可能通过网络获取 NuGet 包。

安装

在 SourceWeft 中

  1. 打开 控制台中的 SharpLens,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

SharpLensMcp

[NuGet] [npm] [License: MIT]

A Model Context Protocol (MCP) server providing 92 AI-optimized tools for .NET/C# semantic code analysis, navigation, refactoring, and code generation using Microsoft Roslyn.

Built for AI coding agents - provides compiler-accurate code understanding that AI cannot infer from reading source files alone.

Installation

Via NuGet (Recommended)

bash
dotnet tool install -g SharpLensMcp

Then run with:

bash
sharplens

Via npm

bash
npx -y sharplens-mcp

Build from Source

bash
dotnet build -c Releasedotnet publish -c Release -o ./publish

Claude Code Setup

  1. Install the tool (pick one):
bash
dotnet tool install -g SharpLensMcp# ornpx -y sharplens-mcp
  1. Create .mcp.json in your project root:
json
{  "mcpServers": {    "sharplens": {      "type": "stdio",      "command": "npx",      "args": ["-y", "sharplens-mcp"],      "env": {        "DOTNET_SOLUTION_PATH": "/path/to/your/Solution.sln (or .slnx)"      }    }  }}
  1. Restart Claude Code to load the MCP server

  2. Verify by asking Claude to run a health check on the Roslyn server

Why Use This with Claude Code?

Claude Code has native LSP support for basic navigation (go-to-definition, find references). SharpLensMcp adds deep semantic analysis:

CapabilityNative LSPSharpLensMcp
Go to definition✅✅
Find references✅✅
Find async methods missing CancellationToken❌✅
Impact analysis (what breaks?)❌✅
Dead code detection❌✅
Complexity metrics❌✅
Safe refactoring with preview❌✅
Batch operations❌✅

Configuration

Environment VariableDescriptionDefault
DOTNET_SOLUTION_PATHPath to a .sln, .slnx or .slnf file (or a directory holding exactly one) to auto-load on startup; the load runs beside the request loop, so initialize answers immediatelyNone (must call load_solution)
SHARPLENS_ABSOLUTE_PATHSUse absolute paths instead of relativefalse (relative paths save tokens)
SHARPLENS_LOG_LEVELLogging verbosity: Trace, Debug, Information, Warning, ErrorInformation
SHARPLENS_TIMEOUT_SECONDSTimeout for long-running operations30
SHARPLENS_MAX_DIAGNOSTICSMaximum diagnostics to return100
SHARPLENS_ENABLE_SEMANTIC_CACHECache the file-path to document lookuptrue (set to false to disable)

The pre-1.6.0 ROSLYN_* spellings of these variables are still read as a fallback; the SHARPLENS_* spelling wins when both are set. Boolean values are case-insensitive. A number that does not parse or is below its minimum falls back to the default, and one above its maximum is capped; each case logs a warning to stderr.

Multi-targeted projects

A project built for several target frameworks is one project to every tool: each result is reported once, with a targetFrameworks list where it exists only under some frameworks. projectName accepts the plain name (MyLib) for all frameworks, or a framework-qualified name (MyLib(net8.0)) for one.

If DOTNET_SOLUTION_PATH is not set, you must call the load_solution tool before using other tools.

Migrating from 1.5.x tool names

Tool names no longer carry the roslyn: prefix — the colon violates the MCP tool-name pattern (^[a-zA-Z0-9_-]{1,64}$), which some clients enforce. Every tool keeps its name minus the prefix:

1.5.x name1.6.0 name
roslyn:load_solutionload_solution
roslyn:get_diagnosticsget_diagnostics
roslyn:rename_symbolrename_symbol
...same rule for all tools...drop the roslyn: prefix

tools/list publishes only the new names. Calls using the old prefixed names are still accepted as aliases for one release and will be removed in the following one.

AI Agent Configuration Tips

AI models may have trained bias toward using their native tools (Grep, Read, LSP) instead of MCP server tools, even when SharpLensMcp provides better capabilities.

To ensure optimal tool usage:

  1. Claude Code: Add to your project's CLAUDE.md:

    For C# code analysis, prefer SharpLensMcp tools over native tools:- Use `search_symbols` instead of Grep for finding symbols- Use `get_method_source` instead of Read for viewing methods- Use `find_references` for semantic (not text) references
  2. Other MCP clients: Configure tool priority in your agent's system prompt

The semantic analysis from Roslyn is more accurate than text-based search, especially for overloaded methods, partial classes, and inheritance hierarchies.

Agent Responsibility: Document Synchronization

Important: SharpLensMcp maintains an in-memory representation of your solution for fast queries. When files are modified externally (via Edit/Write tools), the agent is responsible for synchronizing changes.

When to call sync_documents:

ActionCall sync_documents?
Used Edit tool to modify .cs files✅ Yes
Used Write tool to create new .cs files✅ Yes
Deleted .cs files✅ Yes
Used SharpLensMcp refactoring tools (rename, extract, etc.)❌ No (auto-updated)
Modified .csproj files❌ No (use load_solution instead)

Usage:

# After editing specific filessync_documents(filePaths: ["src/MyClass.cs", "src/MyService.cs"])
# After bulk changes - sync all documentssync_documents()

A full sync also picks up new .cs files the project's compile items include. SharpLensMcp's apply tools write only the files an edit changes, keep each file's encoding and line endings, and refuse with STALE_DOCUMENT when a target file changed on disk since it was read — run sync_documents, then retry. They also refuse with UNSUPPORTED_ENCODING to write a file whose bytes its encoding cannot reproduce (a BOM-less file that is not valid UTF-8 on a system whose fallback code page is UTF-8, such as Linux); save it as UTF-8 first.

Why this design?

This mirrors how LSP (Language Server Protocol) works - the client (editor) notifies the server of changes. This approach:

  • Eliminates race conditions (agent controls timing)
  • Avoids file watcher complexity and platform quirks
  • Is faster than full solution reload
  • Gives agents explicit control over workspace state

If you don't sync: Queries may return stale data (old method signatures, missing new files, etc.)

Features

  • 92 Semantic Analysis Tools - Navigation, refactoring, code generation, diagnostics, discovery, audit/quality
  • AI-Optimized Descriptions - Clear USAGE/OUTPUT/WORKFLOW patterns
  • Structured Responses - Consistent success/error/data format with suggestedNextTools
  • Zero-Based Coordinates - Clear warnings to prevent off-by-one errors
  • Preview Mode - Safe refactoring with preview before apply
  • Batch Operations - Multiple lookups in one call to reduce context usage

Tool Categories

Navigation & Discovery (24 tools)

ToolDescription
get_symbol_infoSemantic info at position
go_to_definitionJump to symbol definition
find_referencesAll references; each classified read/write/invocation/cast/typeof/nameof/attribute; optional kind filter
find_implementationsInterface/abstract implementations
find_callersImpact analysis - who calls this?
get_call_graphMulti-hop callers/callees graph with depth bound + cycle detection
find_path_betweenReachability + the connecting call path(s) between two methods; follows dispatch, with barriers and a checkpoint
get_type_hierarchyInheritance chain
search_symbolsGlob pattern search (*Handler, Get*)
semantic_queryMulti-filter search (async, public, etc.)
get_type_membersAll members by type name
get_type_members_batchMultiple types in one call
get_method_signatureDetailed signature by name
get_derived_typesFind all subclasses
get_base_typesFull inheritance chain
get_attributesList attributes on a symbol
get_containing_memberEnclosing symbol at position
get_method_overloadsAll overloads of a method
find_attribute_usagesFind types/members by attribute
get_external_type_infoInspect NuGet/BCL/external assembly types — members + XML docs
resolve_stack_traceMap a pasted stack trace to file/line/symbol, mangling undone
get_extension_methodsExtensions applying to a type — classic and C# 14 blocks
get_documentationFull XML docs for a symbol with <inheritdoc> expanded
get_super_methodNavigate to the base member / interface members a member implements

Analysis (17 tools)

ToolDescription
get_diagnosticsCompiler errors/warnings + configured analyzer findings (StyleCop, Roslynator, NetAnalyzers); matches CI
diff_api_surfacePublic-API breaking-change report vs a git ref
get_exception_flowWhich exceptions can escape a method, and where they're caught
find_similar_codeStructural similarity search (token-shingle fingerprints)
remove_unused_codeCompute dead-code removals + newly unused usings (generation-only)
find_dead_branchesUnreachable basic blocks per method (real CFG, not heuristics)
add_missing_importsCompute the usings that fix CS0246/CS0103 (generation-only)
analyze_data_flowVariable assignments and usage
analyze_control_flowBranching/reachability
analyze_change_impactWhat breaks if changed?
check_type_compatibilityCan A assign to B?
get_outgoing_callsWhat does this method call?
find_unused_codeDead code detection
validate_codeCompile check without writing
get_complexity_metricsCyclomatic, nesting, LOC, cognitive
find_circular_dependenciesProject and namespace cycle detection
get_missing_membersUnimplemented interface/abstract members

Refactoring (16 tools)

ToolDescription
rename_symbolSafe rename across solution
change_signatureAdd/remove/reorder parameters
extract_methodExtract with data flow analysis
extract_interfaceGenerate interface from class
generate_constructorFrom fields/properties
move_type_to_fileCompute the contents to move a type into its own file (generation-only)
split_typeCompute a partial-class split for selected members (generation-only)
organize_usingsSort and remove unused
organize_usings_batchBatch organize multiple files
format_document_batchBatch format files in project
get_code_actions_at_positionAll Roslyn refactorings at position
apply_code_action_by_titleApply any refactoring by title
implement_missing_membersGenerate interface stubs
encapsulate_fieldField to property
inline_variableInline temp variable
extract_variableExtract expression to variable

Code Generation (3 tools)

ToolDescription
add_null_checksGenerate ArgumentNullException guards
generate_equality_membersEquals/GetHashCode/operators
generate_test_stubCompilable test skeleton for a method (framework auto-detected)

Compound Tools (7 tools)

ToolDescription
get_type_overviewFull type info in one call
analyze_methodSignature + callers + outgoing calls + location
get_file_overviewFile summary with diagnostics
get_method_sourceSource code by name
get_method_source_batchMultiple method sources in one call
get_instantiation_optionsHow to create a type
get_project_healthComposite audit dashboard: diagnostics + unused + coupling + coverage per project

Audit & Quality (10 tools)

ToolDescription
find_god_objectsDetect over-coupled types via efferent + afferent coupling + member-count thresholds
find_untested_codeFind public surface no test reaches — tests detected as xUnit, NUnit, MSTest, TUnit, Reqnroll and SpecFlow discover them, with setup, fixtures and step bindings
find_testsWhich tests cover a symbol — the inverse of find_untested_code
find_type_instantiationsWhere a type is constructed (new T)
find_pattern_usagesWhere a type appears in is/as/pattern matches
find_throw_sitesWhere an exception type is thrown (optionally derived)
find_catch_blocksWhere an exception type is caught (optionally via a base clause)
find_async_issuesasync void / blocking-on-async / unforwarded CancellationToken
check_architectureEnforce namespace/project dependency rules over the type graph
find_naming_violationsNaming audit honoring .editorconfig rules, with conventional defaults

Discovery (3 tools)

ToolDescription
get_di_registrationsScan DI service registrations
find_reflection_usageDetect reflection/dynamic usage
find_interceptorsSurface [InterceptsLocation] call rerouting, generated code included

Infrastructure (12 tools)

ToolDescription
health_checkServer status + load fidelity (partial flag, declared-vs-loaded projects, loadFailures, loadWarnings, lastLoadFailure) — check without reloading
find_unused_dependenciesPackageReferences/ProjectReferences the compiler never needs
fix_allCompute the fix for every instance of a diagnostic id (generation-only)
load_solutionLoad .sln/.slnx/.slnf; reports partial loads (partial + loadFailures: dropped projects, failed restores, projects loaded without references) and MSBuild warnings (loadWarnings) separately
sync_documentsSync file changes into loaded solution
get_project_structureSolution structure
dependency_graphProject dependencies
get_code_fixesAvailable fixes for a diagnostic
apply_code_fixApply a specific code fix
get_nuget_dependenciesNuGet package listing per project
get_source_generatorsList active source generators
get_generated_codeView generated source code

Other MCP Clients

For MCP clients other than Claude Code, add to your configuration:

json
{  "mcpServers": {    "sharplens": {      "command": "sharplens",      "args": [],      "env": {        "DOTNET_SOLUTION_PATH": "/path/to/your/Solution.sln (or .slnx)"      }    }  }}

Usage

  1. Load a solution: Call load_solution with path to .sln or .slnx file (or set DOTNET_SOLUTION_PATH)
  2. Analyze code: Use any of the 92 tools for navigation, analysis, refactoring, audit
  3. Refactor safely: Preview changes before applying with preview: true

Architecture

MCP Client (AI Agent)        | stdin/stdout (JSON-RPC 2.0)        v   SharpLensMcp   - Protocol handling   - 92 AI-optimized tools        |        vMicrosoft.CodeAnalysis (Roslyn)  - MSBuildWorkspace  - SemanticModel  - SymbolFinder

Requirements

  • .NET 8.0 SDK or later — works with .NET 8, 9, 10, and future versions. Analyzes any .NET 8+ project/solution.
  • MCP-compatible AI agent

FAQ

Why does the tool target net8.0 — can it analyze my .NET 9 / .NET 10 project?

Yes. net8.0 is the tool's own runtime floor — the Roslyn 5.x packages it builds on require it — not a ceiling on what it can analyze. RollForward lets the installed tool run on newer .NET runtimes, and MSBuildWorkspace loads each project's real target framework from its csproj, so one install analyzes solutions targeting .NET 8, 9, 10, and beyond.

Development

Adding New Tools

  1. Add the method to the matching src/RoslynService.*.cs partial (Navigation, Analysis, Refactoring, CallAnalysis, …) and return through the shared response envelope:
csharp
public async Task<object> YourToolAsync(string param1, int? param2 = null,    CancellationToken cancellationToken = default){    EnsureSolutionLoaded();    // Your logic...    return CreateSuccessResponse(        data: new { /* results */ },        suggestedNextTools: new[] { "next_tool_hint" }    );}
  1. Register one ToolDefinition in src/ToolRegistry.cs — its name, description, input schema, the ReadOnly flag (mutating tools pass ReadOnly: false and receive a destructiveHint annotation), and a handler that binds arguments through JsonRpcParameters and calls your method. The registry drives both tools/list and dispatch; there is no separate switch to edit. Two tests keep it honest: ToolsListGoldenTests locks the published schema byte-for-byte (re-capture the golden when a schema change is intentional), and ToolSchemaParityTests asserts every parameter the handler reads is declared in the schema.

  2. Build and publish:

bash
dotnet build -c Releasedotnet publish -c Release -o ./publish
  1. Add both test levels — a unit test of the RoslynService method against a deterministic fixture, AND a wire test through the MCP dispatcher (in tests/SharpLensMcp.Tests/Mcp/) with exact value locks plus an error path. This is non-negotiable; see Testing.

Testing

Every test must satisfy the Testing Charter (C1–C9) in tests/SharpLensMcp.Tests/TESTING.md — the standing contract. The headline rules:

  • Lock exact values (C1): assert a concrete name / count / substring / error code / (line, column) — never NotBeNull / > 0 / a type-only check as the sole assertion.
  • Both levels per tool (C4): a unit test against a Fixtures/*.cs fixture and a dispatcher (wire) test that unwraps content[0].text, plus an error path.
  • Right casing (C3): in-process Newtonsoft yields PascalCase error.Code / meta.TotalCount; the MCP wire yields camelCase. Read the casing your test's path actually produces.
  • Deterministic (C7): the suite is serialized via xunit.runner.json; fixture mutators always restore; the timeout test uses a forced-cancellation seam, not a timing race.
  • Out-of-process spine (C6): StdioIntegrationTests value-pins one tool per category over the real binary and runs the tools/list golden over the stdio pipe.
  • Pre-commit gate (C9): build-clean + green is necessary but not sufficient — re-read each changed test and confirm it fails on a wrong answer.

Run the suite:

bash
dotnet test -c Release

Key Files

FilePurpose
src/RoslynService.cs + the src/RoslynService.*.cs partialsTool implementations split by concern across ~30 partials (Navigation, Analysis, Refactoring, Inspection, Validation, TypeDiscovery, Discovery, ExternalApi, Quality, Metrics, CodeActions, CodeGeneration, Compound, CallAnalysis, ExceptionFlow, StackTrace, ApiSurface, SimilarCode, …) — each file's name predicts its contents
src/McpServer.csMCP protocol mechanics: JSON-RPC parse loop, initialize negotiation, per-call timeout, in-band vs protocol error mapping
src/ToolRegistry.cs + src/ToolDefinition.csThe tool surface: one ToolDefinition record per tool (name, schema, ReadOnly flag, handler). Drives tools/list order and dispatch lookup
src/JsonRpcParameters.cs + JsonRpcInvalidParamsException.csTyped JSON-RPC argument accessors and the -32602 Invalid params exception they raise
src/*Data.cs / *Entry.cs records, ConstructorMember.cs, SignatureChange.csTyped records used by the audit composite, constructor generator, and signature-change parser (one type per file)

License

MIT - See LICENSE for details.

来源:README.md,提交 35c4196

工具

0
SourceWeft 尚未在沙箱中启动过该软件包,因此其工具未知。

版本历史

1
  1. v1.5.3最新Sep 16, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分