SwissArmyHammer

SwissArmyHammer is an MCP server that equips AI coding agents with a modular engineering multitool rather than a rigid pipeline. Built in Rust as a single local binary, it connects AI clients such as Claude Code, Cursor, and Windsurf directly to your local development environment, repositories, and shell. Software engineers and development teams use SwissArmyHammer to manage complex implementation tasks without losing context or skipping verification. The server provides a file-backed kanban system, persistent shell execution with searchable history, and AST-level code intelligence via tree-sitter and language server protocols across more than twenty-five programming languages. Instead of enforcing an inflexible sequence of operations, SwissArmyHammer allows agents and developers to plan features, inspect codebases, execute test suites, analyze test coverage, review changes, and generate conventional git commits in any order. Workflows, agents, and custom skills are stored locally as markdown files with frontmatter, allowing teams to version-control custom agent behaviors alongside source code without relying on cloud databases or external hosting services.

Category: AI & LLM Tooling

Tags: file-management, markdown, prompt-engineering, prompts

Visit SwissArmyHammer

How to install and configure SwissArmyHammer

1. Navigate to your project directory in your terminal: cd <your project directory>. 2. Install the command-line package via Homebrew by running brew install swissarmyhammer/tap/swissarmyhammer-cli. 3. Initialize the configuration and server in your repository by executing sah init. 4. Open your MCP client, such as Claude Code, Cursor, or Windsurf. 5. Test the integration and inspect installed workflows by typing /sah-help into your agent chat interface.

What you can do with SwissArmyHammer

• Decomposing feature specifications or conversational requirements into structured kanban tasks, subtasks, and acceptance criteria using the interactive plan skill. * Executing automated test suites and capturing error logs into persistent, searchable shell history without consuming model context window tokens on verbose outputs. * Analyzing codebases across twenty-five languages using tree-sitter symbol lookups, call graph traversals, and blast radius assessments before making refactoring changes. * Automating iterative test-and-fix development cycles using dedicated loops that run code modifications and rerun test suites until all assertions pass. * Conducting structured code reviews that automatically append findings, architectural issues, and missing test coverage items back to the kanban board.

Key facts

• https://github.com/wballard/swissarmyhammer
• AI & LLM Tooling, Files, Documents & PDFs, Notes & Knowledge Management
• file-management, markdown, prompt-engineering, prompts

Part of MCP Servers

How do I install SwissArmyHammer?

You can install the SwissArmyHammer CLI on macOS using Homebrew by running brew install swissarmyhammer/tap/swissarmyhammer-cli. After installation, navigate to your target project folder and run sah init to set up the local MCP server, default skills, and configuration files. You can then access the server tools directly within your MCP-compatible client.

What can SwissArmyHammer do?

SwissArmyHammer equips coding agents with a full suite of software development tools. It provides file-backed kanban planning, code implementation, test execution, test coverage analysis, AST-based symbol exploration, structured code review, and git commit generation. It also includes a smart shell with searchable output history and automatic codebase indexing using tree-sitter and language servers across more than twenty-five languages.

Which MCP clients work with SwissArmyHammer?

SwissArmyHammer works with any client supporting the Model Context Protocol. Supported environments specifically include Claude Code, Cursor, and Windsurf, along with custom agent implementations that connect to local MCP servers. Because the server runs as a local native binary, it integrates directly with your local shell, file system, and git configuration.

How does SwissArmyHammer manage context window limits?

SwissArmyHammer manages context limits through a smart virtual shell and AST-level code parsing. Command execution outputs are indexed and stored locally, returning truncated summaries to the model while allowing regex grep, semantic search, and line-range retrieval on full outputs. Additionally, tree-sitter indexing enables targeted symbol lookups and call graph analysis without dumping entire files into the agent prompt.