An enhanced MCP server that enables AI assistants to execute Claude Code with bypassed permissions, task orchestration, and robust error handling.
Claude Code MCP Enhanced is a Model Context Protocol (MCP) server that allows AI assistants like Claude Desktop to run Claude Code with bypassed permissions and advanced orchestration capabilities. It solves the problem of AI assistants struggling with multi-step edits by providing a unified tool for executing complex coding tasks, file operations, and system commands directly through the Claude CLI. The server enhances reliability with features like automatic retries, heartbeat mechanisms, and task breakdown patterns.
Developers and teams using AI coding assistants (like Claude Desktop, Cursor, or Windsurf) who need to automate complex, multi-step coding tasks, refactor codebases, or execute system operations directly through their AI agent.
Developers choose this over standard Claude Code integration because it adds orchestration, reliability, and automation features—like the boomerang pattern for task decomposition and markdown-to-command conversion—making AI-driven development more robust and scalable. It turns Claude Code into a programmable agent capable of handling intricate workflows without manual intervention.
Enhanced Claude Code MCP server with orchestration capabilities, reliability improvements, and self-contained execution patterns
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Provides a single `claude_code` tool that executes prompts via Claude CLI with the `--dangerously-skip-permissions` flag, enabling direct file edits and system operations without interruption, as highlighted in the Key Features.
Implements the boomerang pattern to break complex workflows into manageable subtasks, allowing parent tasks to delegate and aggregate results, which is essential for multi-step coding operations described in the Task Orchestration section.
Includes a `convert_task_markdown` tool that automatically transforms human-readable task lists into executable MCP commands, bridging planning and execution as demonstrated in the Task Converter workflow.
Features heartbeat mechanisms, automatic retries, and graceful shutdown to prevent timeouts and improve robustness, with configurable environment variables like `MCP_HEARTBEAT_INTERVAL_MS` for long-running tasks.
Supports Roo modes integration via a `.roomodes` configuration file, enabling tailored task execution with roles like coder or designer, as explained in the Roo Modes Integration section.
Requires a manual first-time step to accept the `--dangerously-skip-permissions` flag in Claude CLI and precise MCP client configuration, which can be error-prone and time-consuming, as noted in the Important First-Time Setup section.
Automatically bypasses permissions, posing potential risks in untrusted environments due to the `--dangerously-skip-permissions` flag, a concession acknowledged in the README's warnings.
Heavily relies on Claude CLI being correctly installed and maintained; issues like command not found or version mismatches can break functionality, as mentioned in the Troubleshooting section.
Primarily designed for Claude-based AI assistants and MCP clients, making it less effective for teams using other AI coding tools without MCP integration, limiting its versatility.