Shipping an MCP server means nothing if developers can't find it. The three places they actually look — the official MCP Registry, Smithery, and GitHub Topics — each require a separate submission, and the MCP Registry is what Glama and PulseMCP pull from automatically, so skipping it means skipping those too.
marketing onboard --name my-tool --repo owner/repo --kind mcp-server handles all three in one step. It fetches your README, extracts the problem and key facts via Claude, writes them to projects.yml, and files the directory submissions. After that, a GitHub Actions cron at 14:00 UTC on weekdays rotates through four social channels — Bluesky, Mastodon, Dev.to, Hashnode — picking the least-recently-used angle each run.
The part I spent the most time on is the antislop gate in pipeline/antislop.py. It hard-rejects specific tokens before any post goes out: excited, game-changer, unlock, empower, AI-powered, emoji, hashtags, exclamation points, and rhetorical questions. Every post has to clear that filter or it doesn't ship. Per-channel length limits are enforced the same way — 300 chars for Bluesky, 280 for X, 500 for Mastodon, 150–400 words for Dev.to and Hashnode.
One caveat worth knowing: awesome-claude-code submission can't be automated. Their rules require a human to file via their GitHub issue form. The pipeline generates the payload, but you submit it once manually.
You need an Anthropic API key plus credentials for at least Bluesky, Dev.to, and Hashnode to run the full pipeline. Mastodon and Slack are optional.
Top comments (0)