Helmdeck, developed by Tosin2013, is an MCP server built with Python 3.12+ for AI-driven development. It integrates various Large Language Models, from local models like Gemma 4 to frontier models such as Claude. Designed to enhance developer productivity, Helmdeck connects to external tools and data sources via the MCP standard and manages the Helm Kubernetes package manager. It supports agentic workflows, providing specialized capabilities for code analysis, architectural decision-making, and feature implementation. Available as a native macOS application or via the MatrixHub CLI, Helmdeck serves as a central hub for AI-powered software engineering tasks.

What does Helmdeck connect to?

Helmdeck functions as an MCP server, linking to diverse Large Language Models. It supports local or efficient LLMs like gpt-oss-120b, Gemma 4, and Mistral, alongside advanced models such as Claude, Codex, and Qwen. The server also integrates with external tools and data sources through the MCP standard. Furthermore, Helmdeck provides direct control over the Helm Kubernetes package manager, extending its utility for deployment and management tasks. This broad connectivity allows it to serve as a versatile platform for various AI-powered development needs.

Core features and tools

Helmdeck offers 36 typed capability packs for browser, code, slides, vision, and desktop workflows. Key callable tools include AI-powered architectural decision analysis and ADR management, semantic code understanding, extracting function signatures, identifying architectural patterns, and AI-powered discovery of code files related to ADRs and architectural decisions. For Helm integration, it provides Discovery & search, Installation & lifecycle, Validation & safety, Monitoring & status, Multi-cluster, and Built-in guidance capabilities.

Who is it for?

Helmdeck is designed for developers who ship with AI and development teams. It targets users aiming to run agentic workflows on local or efficient Large Language Models. This includes individuals and teams focused on tasks like codebase migration, new feature addition, and AI-powered architectural decision analysis. The server supports a range of development activities, from processing design assets to generating integration tests. Skip this if your primary need is a simple, standalone LLM interface without deeper integration into development workflows or Kubernetes management.

How to get started

No credentials are required for direct client connection to the local HTTP endpoint. Optional integrations use API key-based authentication, such as FIRECRAWL_API_KEY, with credentials managed internally by vault-backed systems.

Use cases

• Migrating a codebase, such as from REST to tRPC, by analyzing routes, planning the migration, generating new code, and validating type contracts.
• Adding new features, like Supabase login with magic links, through processing design assets, scaffolding authentication clients, building UI components, generating email templates, and writing integration tests.
• AI-powered architectural decision analysis and management, involving semantic code understanding, identifying architectural patterns, and automating development workflows.

Helmdeck provides a specialized MCP server for AI-powered development, integrating various Large Language Models and offering distinct capabilities for code analysis and architectural management. It supports complex agentic workflows, making it a valuable asset for development teams. The server's ability to manage Helm Kubernetes and connect to external services enhances its utility for modern software engineering.