Autocontext by 2site Net is a free Visual Studio Code extension designed as an MCP server tool. Built with TypeScript, it enhances AI coding assistants, particularly Copilot, by injecting context and performing quality checks on codebases. It operates within the Model Context Protocol, which standardizes how servers provide argument autocompletion suggestions for prompts and resource URIs to AI tools. This extension targets developers seeking to enrich their AI assistant's understanding of project-specific guidelines and ensure code quality directly within their development environment.

What does Autocontext connect to?

Autocontext connects directly to AI coding assistants, notably integrating with Copilot's context. As a Model Context Protocol (MCP) server tool, it uses this standardized framework to deliver contextual information and quality evaluations for a codebase. The MCP defines how servers offer argument autocompletion suggestions for prompts and resource URIs to various AI tools. This connection allows Autocontext to extend the capabilities of AI assistants by providing them with deeper, project-specific insights during development tasks within the Visual Studio Code environment.

Core features and tools

Autocontext provides Language Model tools specifically for AI assistants. It exposes capabilities such as list_autocontext_instructions_files for enumerating available instruction sets and search_autocontext_instructions_files_by_metadata for finding relevant guidelines based on specific criteria. These tools allow AI agents to dynamically discover and apply appropriate instructions. Beyond these named functions, the extension performs MCP Tool Checks, ensuring quality assurance across different aspects of a codebase, including coding style, dependency management, commit formatting, and configuration consistency.

Who is it for?

Autocontext is for developers utilizing AI coding assistants who need project-specific context and quality adherence. It supports workflows injecting curated Markdown guidelines, or Chat Instructions, into the AI assistant's context. AI clients like Copilot can perform MCP Tool Checks in Agent mode, covering .NET coding style, NuGet hygiene, Git commit format, EditorConfig property resolution, and TypeScript coding style. Skip this if your development workflow does not involve AI coding assistants.

How to get started

No credentials are required. The tool is distributed as a Visual Studio Code extension, installed via the VS Code Quick Open command (Ctrl+P).

Use cases

• Providing curated Markdown guidelines ("Chat Instructions") to AI assistants for various programming languages and frameworks.
• Performing MCP Tool Checks for quality assurance on aspects like .NET coding style, NuGet hygiene, and Git commit format.
• Enabling AI agents to discover relevant instructions by listing and searching available instruction files.

Autocontext by 2site Net provides a valuable bridge between codebases and AI coding assistants within Visual Studio Code. It enhances AI capabilities by offering contextual instructions and enforcing quality checks through its MCP server tool features. Built with TypeScript, this free extension helps developers maintain consistent code standards and guides AI interactions with project-specific knowledge. Try performing MCP Tool Checks for quality assurance on your .NET coding style.