Klavis Documentation

The Poste.io Local MCP server is a FastMCP-based server designed for email management via IMAP/SMTP through the Model Context Protocol (MCP). It allows for secure, isolated email operations on your local machine.

Capabilities include:

• Email Management: Get, read, and search emails with pagination support.
• Sending/Replying: Send and reply to emails, including attachment handling.
• Folder Operations: List, create, and delete email folders.
• Draft Management: Save, update, and delete email drafts.
• Backups: Import and export email backups.
• Mailbox Insights: Access mailbox statistics.
• Security: Provides workspace isolation and security for local operations.

The Notion MCP Server implements the Model Context Protocol (MCP) for the Notion API. It provides tools specifically designed for AI agents, such as editing pages in Markdown with optimized token consumption.

Note on Support: Notion is prioritizing Notion MCP (remote), which uses standard OAuth for easy installation. This local MCP server repository may be sunset in the future and is not actively monitored for issues. For support regarding the remote version, contact Notion support directly.

The GitHub MCP Server connects AI tools (agents, assistants, chatbots) directly to GitHub. It enables natural language interactions for:

• Repository Management: Browsing code, searching files, analyzing commits, and understanding project structure.
• Issue & PR Automation: Creating, updating, and managing issues and pull requests.
• CI/CD & Workflow Intelligence: Monitoring GitHub Actions, analyzing build failures, and managing releases.
• Code Analysis: Examining security findings, reviewing Dependabot alerts, and understanding code patterns.
• Team Collaboration: Accessing discussions, managing notifications, and analyzing team activity.

The Excel (Local) MCP server allows you to create, read, and modify Excel workbooks directly on your machine without requiring Microsoft Excel to be installed. It supports advanced Excel features including formulas, formatting, charts, pivot tables, and Excel tables.

Key Capabilities:

• Workbook Operations: Create, read, and update workbooks and worksheets.
• Data Manipulation: Use formulas, create charts, and manage pivot tables.
• Formatting: Apply rich formatting such as fonts, colors, borders, and alignment.
• Table Management: Perform table operations with custom styling.
• Data Integrity: Perform data validation and integrity checks.
• Transport Support: Compatible with stdio, SSE, and streamable HTTP transports.

Postgres MCP Pro is designed to complement generative AI by providing deterministic tools and classical optimization algorithms rather than relying solely on LLM reasoning.

Key differentiators include:

• Deterministic health checks: Provides repeatable results compared to LLM-generated queries.
• Principled indexing: Uses established search strategies instead of Gen-AI guesses.
• Workload analysis: Identifies top problems via actual data rather than inconsistent analysis.
• Performance simulation: Simulates improvements rather than requiring manual verification.

It incorporates health checks from PgHero and implements modern industrial-strength algorithms for automatic index tuning.

The GitHub MCP Server supports two deployment modes:

1. Local GitHub MCP Server: Runs locally alongside your IDE or application. It requires Personal Access Tokens (PATs) for authentication. It is compatible with both GitHub Enterprise Server (GHES) and GitHub Enterprise Cloud (GHEC).

2. Remote GitHub MCP Server: Runs as a hosted service accessed over the internet. It is currently available only on GitHub Enterprise Cloud (GHEC). It supports authentication via GitHub App Installation Tokens, OAuth 2.0 Authorization Code flow, or Personal Access Tokens (PATs).

The Word (Local) MCP server allows AI assistants to interact with Microsoft Word documents on your local machine. It provides a standardized interface for creating, reading, and editing documents with rich formatting capabilities.

Key Capabilities:

• Document Management: Create, read, copy, and merge documents.
• Content Creation: Add headings, paragraphs, tables, images, and lists.
• Rich Text & Styling: Apply bold, italic, underline, and colors; format and style tables.
• Advanced Features: Add footnotes and endnotes, perform search and replace, manage document protection/properties, and convert documents to PDF.

The Terminal (Local) MCP server allows LLM applications to execute controlled command-line operations on your local machine. It is designed with security-first features to prevent unauthorized access and malicious commands.

Key Security Features:

• Whitelisting: Configurable command and flag whitelisting to restrict allowed operations.
• Injection Protection: Prevents shell operator injection and path traversal attacks.
• Execution Controls: Supports execution timeouts, length limits, and working directory restrictions.
• Validation: Secure command execution with detailed error reporting and async support.

The Filesystem (Local) MCP server is a Node.js-based server that implements the Model Context Protocol (MCP) to allow LLMs to perform local filesystem operations. It provides capabilities for managing files and directories with dynamic access control enforced via the Roots protocol.

Key Capabilities:

• File Management: Read and write files.
• Directory Management: Create, list, and delete directories.
• File Operations: Move files and directories.
• Search: Find files using glob patterns.
• Inspection: Retrieve file metadata and directory trees.
• Security: Dynamic directory access control via Roots.