v0.1.6 - Critical Serialization Bugfix

Fixed

Critical bugfix: Read-only tools now return actual data in TextContent.text.

Previously, the following tools returned only static messages instead of actual JSON data:

• list_directory - was returning str(items)
• get_document_structure - was returning "Structure extracted"
• search_text - was returning "Found X matches"
• get_context - was returning "Context extracted"
• read_element - was returning "Element read"

This meant that LLM agents could not "see" the results of these queries, making them effectively blind when navigating documents.

All read-only tools now return proper JSON-serialized data that agents can parse and use.

---

Full Changelog: v0.1.5...v0.1.6

v0.1.5 - Smithery Configuration Improvements

Changed

• Removed incompatible Smithery Build GitHub Actions workflow (container runtime projects deploy via Smithery web interface, not CLI)
• Updated smithery.yaml to modern container runtime format
• Added .dockerignore for optimized Docker builds

Fixed

• Clarified Smithery deployment method in documentation (web-based deployment for container runtime)

Deployment

This version is published to PyPI and can be deployed to Smithery via their web interface at https://smithery.ai/new

v0.1.4 - Smithery Cloud Integration

This release completes the integration with Smithery.ai, adding automated builds and configuration for cloud deployment.

v0.1.3 - PyPI SEO & Visibility

This release adds metadata (keywords, classifiers, and project URLs) to improve the project's discoverability on PyPI and search engines.

v0.1.2 - Windows Compatibility & Ollama Tests

Fixed

• Fixed Windows compatibility issues by making fcntl usage optional.
• Fixed critical bug in Markdown parser where heading nesting was incorrect.

Added

• Comprehensive integration tests with Ollama (tests/test_ollama_tools.py).
• Interactive CLI editor script (interactive_editor.py).

v0.1.0 - Initial Release

Initial Release

Added

• Initial release of the Markdown Editor MCP Server.
• Semantic editing tools: get_document_structure, read_element, replace_content, insert_element, delete_element.
• Narrative flow tools: get_context, move_element.
• Information retrieval: search_text.
• Metadata management: YAML Frontmatter support via update_metadata.
• File system operations: list_directory, create_file, create_directory, delete_item.
• Robust Markdown parser with structural element tracking.
• Transaction support (internal logic) and change journaling.

Features

• 15 MCP Tools for comprehensive Markdown editing
• MCP 2025 Standard Compliance - Tool Use Examples and Output Schemas for all tools
• Semantic Navigation - Structure-aware editing and navigation
• YAML Frontmatter - Metadata management support
• Change History - Transaction support with undo capability
• GitHub Actions - Automated PyPI publishing on release

Installation

pip install markdown-editor-mcp-server

Documentation

See README.md for full documentation.