Scheduling social posts from inside Claude Code — anyone else building MCP tools as distribution?

Something I've been thinking about while building XreplyAI: the people who are most likely to care about staying consistent on social are the same people spending 8 hours a day in their editor. The context switch to a browser dashboard is exactly where the habit breaks.

So I shipped an MCP server for XreplyAI. One config block in .mcp.json and Claude Code gets tools to draft, schedule, and publish posts across X, LinkedIn, Instagram, Threads, YouTube, Pinterest, Bluesky, and TikTok — without leaving the session.

The workflow that clicked for me: finish a PR, ask Claude to "draft a building-in-public post about what I just shipped in my voice," review it, schedule it. The voice piece matters — XreplyAI trains on your tweet archive so the output actually sounds like you rather than generic AI copy.

What I'm curious about: is anyone else treating MCP servers as a distribution channel? It feels early but the install friction is low (just npx + a config line) and the audience — people deep in AI-native workflows — seems like exactly where early SaaS adoption happens right now.

How do you handle the "I https://xreplyai.com?utm_source=producthunt&utm_medium=social&utm_campaign=feature-2026-06-01 about this" moment when you're heads-down building?