Feature/mcp

[adeljck]

feat: add MCP server mode for AI integration

Description

What will it do?

This PR introduces native support for the Model Context Protocol (MCP), enabling AI agents (like Cursor, Claude Desktop, etc.) to invoke dirsearch scans directly via STDIO or HTTP transport.

Key implementation details:

• MCP Module: Implemented FastMCP server exposing dirsearch_scan and dirsearch_scan_cli tools.
• Security: Added Bearer Token authentication (--mcp-token) with auto-generation support for HTTP mode.
• Packaging: Updated PyInstaller specs and setup.py (pip install ".[mcp]") to include fastmcp dependencies.
• Docker: Added Dockerfile.mcp for building MCP-enabled container images.
• Docs & Tests: Updated README with client configuration examples and added 9 unit tests for MCP logic.

Requirements

---

• Add your name to CONTRIBUTORS.md
• If this is a new feature, then please add some additional information about it to CHANGELOG.md
• All tests pass (python -m unittest tests)
• Documentation is updated (README.md)

How to Use

1. Install Dependencies:

pip install -r requirements/mcp.txt

2. Run MCP Server:

# STDIO Mode (Local)
python3 dirsearch.py --mcp

# HTTP Mode (with Auth)
python3 dirsearch.py --mcp --mcp-transport http --mcp-token

# if leave mcp-token value empty,system will generate a token randomly.

Pics