My usual distribution workflow: write a Bluesky post (300 chars), trim it for X (280), expand it for Mastodon (500), write something longer for Dev.to, check what angle I used last time, forget, repeat a week later. That's before touching directory submissions.
I built marketing-pipeline to collapse that into a single configured-once system. The core mechanic: every project lives in projects.yml with a list of angles — specific framings of the problem the tool solves. The daily cron (marketing cycle, running at 14:00 UTC via GitHub Actions) picks the least-recently-used angle per project, generates posts for each channel, enforces the length constraints automatically (280/300/500 chars for micro-platforms, 150–400 words for Dev.to and Hashnode), and publishes.
Onboarding a new project is one command:
marketing onboard --name my-tool --repo owner/repo --kind mcp-server
That fetches the README, sends it to Claude, and writes the problem statement, facts, and angles to projects.yml. From there the daily rotation handles the rest.
One piece I spent real time on: the antislop gate in pipeline/antislop.py. Before any post publishes, it hard-rejects a specific token list — 'excited', 'game-changer', 'unlock', 'AI-powered', emoji, hashtags, exclamation points, rhetorical questions. The gate runs before posting, not after. A draft that hits any of those tokens gets thrown out and regenerated. Growth-marketing language is the failure mode for developer tool posts, and I wanted that enforced mechanically, not by memory.
The kind field routes projects to the right directories automatically. mcp-server hits MCP Registry, Smithery, Glama, and PulseMCP. claude-skill targets awesome-claude-code (though that one requires a human to submit via their GitHub issue form — the pipeline generates the payload, you click submit). browser-extension covers Chrome Web Store, Firefox AMO, and Edge Add-ons.
Requirements: Anthropic API key plus credentials for Bluesky, Dev.to, and Hashnode at minimum. Mastodon and Slack are optional add-ons.