The lsp-mcp server provides a unified API for multi-language code analysis.
Analyze Files: Run static code analysis on files by specifying the path via the AnalyzeFile endpoint.
Multi-Language Support: Currently supports Python (via Ruff) with plans to add Go, Rust, and more.
Automated Sidecar Management: Manages the langtools_daemon sidecar that executes linters and analysis tools.
Integration-Friendly: Process analysis requests in a headless and batch-friendly manner, suitable for automation, CI/CD pipelines, and integration with LLMs.
JSON Output: Returns normalized analysis results in JSON format for easy integration with other tools.
Integration for code repository access and analysis, supporting Git-based projects
Repository management integration, allowing code analysis for GitHub-hosted projects
Integration with Python Package Index for dependency management and installation
Comprehensive Python language analysis, error detection, and code quality checking
Provides fast Python code analysis, linting, and static checking through integration with Ruff
Planned integration with rust-analyzer for Rust code analysis and static checking (mentioned in roadmap)
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@lsp-mcp lint this Python function for style issues and unused imports"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
🚧 This is actively being developed, so expect issues. Currently focusing on compatibility with the on-machine AI agentGoose. 🚧
langtools-mcp is a Model Context Protocol (MCP) server and client toolkit that gives LLMs and AI agents unified access to real static analysis tools—including batch CLI checkers (like Ruff and go vet) and LSPs (like gopls, rust-analyzer, and more).
We've all been there:
LLM writes code that doesn't compile due to hallucinating standard libraries that don't exist, syntax errors, etc.
LLM writes functional code, but the style, formatting, and linting is non-existent (I'm looking at you, unused imports)
langtools-mcp aims to help solve this by letting your AI and agentic apps catch, explain, and even fix issues in code, by calling the same tools expert programmers use. The goal is simply for this MCP to be a tool that the LLM begins using as part of its dev cycle. Just like IDEs and LSPs supercharged humans' ability to quickly assess and fix issues during the dev process, langtools aims to do this with an MCP tool.
🧠Supercharge Agents: Let your LLMs/AI validate, lint, and debug their own code.
🧩 Modular & Extensible: Add new languages/tools in minutes via strategies.
âš¡ Daemon or Batch: Runs as a fast HTTP daemon for LSP and batch CLI tools.
langtools:
args:
- --from
- https://github.com/flothjl/langtools-mcp
- langtools-mcp
bundled: null
cmd: uvx
description: null
enabled: true
env_keys: []
envs: {}
name: langtools-mcp
timeout: null
type: stdioexport LANGTOOLS_PYTHON_TOOLS='["ruff"]'
export LANGTOOLS_GO_GO_TOOLS='["vet"]'Language | Tools |
Python | ruff, pyright |
Go | vet |
Related MCP server: MCP Code Analysis Server
git clone https://github.com/flothjl/langtools-mcp.git
cd langtools-mcp
uv sync # or pip install -e .[dev]Requirements: Python 3.10+, plus ruff, pyright, and Go for Go support (must be in your PATH).
Python: Ruff, Pyright (CLI)
Go: go vet (CLI)
Rust: rust-analyzer (LSP)
JavaScript/TypeScript: tsc, eslint (planned)
Want to add support for your favorite tool or language? Open a PR or start a Discussion!
Fork, clone, and submit a PR!
Code and docs welcome for new languages, better error messages, and more.
Run a file through analysis. You MUST provide a valid Path for the file.
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It mentions the need for a 'valid Path', implying some validation, but doesn't disclose behavioral traits like what the analysis entails, whether it's read-only or destructive, error handling, or output format. The description is minimal and leaves key behaviors unspecified.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two short sentences with zero waste—it directly states the action and a critical requirement. It's appropriately sized and front-loaded, making it easy to parse quickly without unnecessary details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (analysis implies some processing), lack of annotations, no output schema, and low schema coverage, the description is incomplete. It doesn't explain what 'analysis' means, what results to expect, or any behavioral context, making it inadequate for an agent to understand the tool's full use.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema has 1 parameter with 0% description coverage, so the description must compensate. It adds meaning by specifying that the 'file_path' must be a 'valid Path', clarifying the parameter's purpose beyond the schema's basic 'File Path' title. However, it doesn't detail path formats or constraints, leaving some gaps.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states the action ('Run a file through analysis') and resource ('file'), but 'analysis' is vague—it doesn't specify what type of analysis (e.g., security, content, format) or what the analysis produces. Without sibling tools, differentiation isn't needed, but the purpose lacks specificity beyond a generic verb.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description includes a mandatory requirement ('You MUST provide a valid Path'), which is a basic usage constraint, but it doesn't explain when to use this tool (e.g., for what scenarios or goals) or any prerequisites beyond the path. No alternatives are mentioned, but since there are no sibling tools, this is less critical.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
With only one tool, there is no possibility of ambiguity or overlap between tools. The tool's purpose is clearly defined as analyzing a file, leaving no room for confusion or misselection.
A single tool inherently has perfect naming consistency, as there are no other tools to compare against. The tool name 'AnalyzeFile' follows a clear verb_noun pattern, which would be consistent if more tools were added.
One tool is too few for a server named 'lsp-mcp', which suggests a Language Server Protocol domain. A single analysis tool lacks the expected scope for language features like diagnostics, completions, or refactoring, making it feel incomplete and thin.
The tool set is severely incomplete for an LSP-related server. It only provides file analysis, missing essential operations such as getting diagnostics, providing code completions, or handling references, which are core to language server functionality.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Code intelligence for LLMs. Analyze, search, and retrieve code from any public git repository.
Enterprise code intelligence for M&A, security audits, and tech debt. Hosted server with 200k free.
Lints + auto-fixes how AI coding agents discover any new product. 24 rules, 6 tools, score 0-100.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/flothjl/langtools-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
