What is Agenticmcp

AgenticMCP (Agentic Model Communication Protocol) is an open-source TypeScript-based command-line interface (CLI) tool that allows users to utilize various Large Language Models (LLMs) as predefined agents with specific roles and contexts.

Use cases

Use cases for AgenticMCP include generating reports, writing code, summarizing documents, and serving as an MCP server for tool-based AI workflows.

How to use

To use AgenticMCP, install it via npm with ‘npm install -g agenticmcp’. You can list available commands with ‘agenticmcp --help’ and run agents with specific contexts, such as ‘agenticmcp writer requirements.txt’ or ‘agenticmcp serve:mcp --provider openai’ to start an MCP server.

Key features

Key features of AgenticMCP include a unified CLI for multiple LLM providers, predefined agent commands with role-based context, secure credential management, a tool system for file/code operations, modular architecture, and cross-platform compatibility.

Where to use

AgenticMCP can be used in various fields such as software development, data analysis, content creation, and any domain requiring interaction with large language models for specific tasks.

AgenticMCP CLI

Overview

AgenticMCP (Agentic Model Communication Protocol) is an open source TypeScript-based command-line interface (CLI) tool that enables users to leverage various Large Language Models (LLMs) as predefined agents with specific roles and contexts. It provides a unified, extensible, and secure interface for interacting with LLM providers such as OpenAI, Anthropic, Google, and more.

Features

• Unified CLI for multiple LLM providers
• Predefined agent commands with role-based context
• Secure credential management (keytar integration)
• Tool system for file/code operations and shell integration
• Modular, testable architecture with dependency injection
• Cross-platform (Linux, macOS, Windows)

Installation

npm install -g agenticmcp

Usage

# List available agent commands
agenticmcp --help

# Run a writer agent with file context
agenticmcp writer requirements.txt

# Use multiple files or directories as context
agenticmcp analyst report1.pdf data.csv
agenticmcp summarizer ./docs/

# Start as an MCP server for tool-based AI workflows
agenticmcp serve:mcp --provider openai

See REQUIREMENTS.md for detailed CLI examples and advanced usage.

Architecture

• Core Framework: Configuration, credentials, type system, and logging
• Provider System: Factory pattern for LLM providers (OpenAI, Anthropic, Google, Grok)
• Command System: Decorator-based registry for agent commands
• Context Management: File/directory context ingestion, token optimization
• Tool System: Registry and executor for tool calls, shell/file operations
• Conversation Management: Multi-turn session state, tool call integration

See KNOWLEDGE.md for a detailed architecture analysis.

Security

• Credentials stored securely via keytar in the OS keychain
• API keys resolved from env, config, or keychain (see REQUIREMENTS.md)
• Data protection, error handling, and privacy controls

Contributing

Contributions are welcome! Please see CONTRIBUTING.md if available, or open an issue/pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.