🔌
Universal Integration
Convert any command-line tool into an MCP-compliant agentic toolset through a standards-based framework.
CoalaCommand-line LLM-agent Adapter
A standard-based framework for converting command-line tools into agentic toolsets
Skip to content
Overview
Coala (Command-line LLM-agent Adapter) is a standards-based framework that converts command-line tools into agentic toolsets by bridging the Model Context Protocol (MCP) and Common Workflow Language (CWL). Coala treats tool definitions as data rather than hard-coded logic, allowing Large Language Models (LLMs) to discover and execute reproducible analyses without the need to write custom wrappers for every individual tool.
- MCP acts as a "USB-C port", standardizing how LLMs connect to tools and data, regardless of the underlying infrastructure or vendor.
- CWL acts as the "Blueprint," providing an open standard for describing command-line tools with validated schemas and containerized environments.
How It Works
The Coala framework, implemented in a Python package, operates on a three-tier architecture to translate natural-language queries into reproducible tool execution.
- Client Layer: Any MCP-compliant client application (e.g., Claude Desktop, Cursor, or custom interfaces) that utilizes LLMs (such as Gemini, GPT-5, or Claude) to enable natural language interaction.
- Bridge Layer: A local, generic MCP server that acts as a schema translator. Unlike standard MCP servers that require custom Python wrappers for each tool, the bridge layer automatically parses CWL definitions and exposes the CWL-described command-line tools as executable MCP utilities.
- Execution Layer: A standard CWL runner that executes the underlying binaries within containerized environments (Docker). This ensures that analyses are reproducible and isolated from the host system's dependencies.
The user sends a natural language query to the MCP Client (e.g., Claude Desktop). The Client retrieves the tool list from the MCP server. The LLM selects the appropriate tool and sends a structured request for the analysis. Coala translates this selection into a CWL job and executes it within a container (Docker), ensuring reproducibility. The execution logs and results are returned to the LLM, which interprets them and presents the final answer to the user.