Slash Command Manager provides an MCP server that exposes a programmatic API for generating and managing slash commands as part of the Spec-Driven Development workflow. You can run the MCP server in HTTP mode to accept requests from MCP clients and automation tooling, enabling integration with CI pipelines and other services that need to generate slash commands on demand. The repository also includes a standalone CLI (slash-man) for interactive generation and management of slash command definitions, as well as features for code pattern detection and prompt-based generation across multiple AI assistants. Use the MCP endpoint to programmatically request new or updated slash command definitions, or leverage the CLI for manual generation and exploration before exposing results through the server.

Prerequisites:

• Python 3.8+ installed on your system
• git available on your PATH
• Optional: uv (the package manager) if you want to install via uvx before publishing

Install from source:

1. Clone the repository: git clone https://github.com/liatrio-labs/slash-command-manager.git
2. Install in editable mode (from repository root): uv pip install -e .

   or if using pip directly:

   python -m pip install -e .

Run MCP server locally (HTTP transport):

1. Start the MCP server using the CLI: slash-man mcp --transport http --port 8000 2)(Optional) Provide a custom configuration file: slash-man mcp --config path/to/config.toml --transport http --port 8080

If you prefer using uvx (when published): uvx --from git+https://github.com/liatrio-labs/slash-command-manager slash-man mcp --transport http --port 8000

Tips and considerations:

• The MCP server can be run with HTTP transport to integrate with automation tooling; the default port in examples is 8000.
• Environment variable CONFIG_FILE can be used to point the server at a custom TOML/JSON configuration, if supported by the build.
• The Slas-Command-Manager project includes a CLI (slash-man) for interactive generation of slash commands and for testing via the provided commands; use the CLI to bootstrap prompts and command structures before exposing them via MCP.
• If you are debugging, you can use the --help flags on the CLI to discover available subcommands (generate, cleanup, mcp) and options.
• Versioning follows a format VERSION+COMMIT_SHA when available, which helps track builds against git history.
• For GitHub prompts and repository-driven prompts, ensure you have network access and proper permissions; some features rely on public GitHub repos as shown in the Quick Start prompts.