Publish lorehive to PyPI and MCP directories

[itdove]

Summary

After MVP completion, publish to PyPI and MCP directories for installation and discovery.

PyPI package name: lorehive (both lore and lore-mcp are taken on PyPI)

• Repo stays lore, CLI command stays lore
• Install: pip install lorehive
• Use: lore init, lore sync, lore search

1. PyPI Publication

• Update pyproject.toml: name = "lorehive"
• Publish lorehive package to PyPI
• Set up GitHub Actions workflow for automated release on tag push
• Ensure pyproject.toml metadata is complete (description, URLs, classifiers, keywords)
• Add PyPI badge to README

2. MCP Directory Submissions

Submit to major MCP directories for discoverability:

Directory	URL	How to submit
mcp.so	https://mcp.so	PR or submission form
awesome-mcp-servers	https://github.com/punkpeye/awesome-mcp-servers	PR to GitHub repo
Smithery	https://smithery.ai	Register package
Glama	https://glama.ai/mcp/servers	Submit via site
Claude Code registry	Built-in to Claude Code	Anthropic submission (when available)

3. Prerequisites

• README with clear install instructions, usage examples, and screenshots/demos
• Stable API (no breaking changes expected short-term)
• CI/CD passing on main
• License file in place (Apache-2.0 already set)
• CHANGELOG with release notes

Acceptance Criteria

• pyproject.toml updated: name = "lorehive"
• pip install lorehive works from PyPI
• CLI command remains lore (via [project.scripts])
• GitHub Actions publishes on version tag push
• Listed on at least mcp.so and awesome-mcp-servers
• README has install instructions, quick start, and feature overview
• Smithery listing created (if platform supports pip packages)