The latest articles on Forem by Omnivocal Dev (@omnivocal). https://forem.com/omnivocal https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3838238%2Fbd5cbf9a-cc87-41c5-8baf-7fec95e141d4.png https://forem.com/omnivocal en Omnivocal Dev Sun, 05 Apr 2026 19:33:11 +0000 https://forem.com/omnivocal/dev-log-2-omnivocal-in-production-127p https://forem.com/omnivocal/dev-log-2-omnivocal-in-production-127p <h1> Dev Log #2: OmniVocal in Production </h1> <p>Two weeks since Dev Log #1. Here's what shipped.</p> <h2> 10 channels now live </h2> <p>Bluesky, dev.to, Hashnode, GitHub, Mastodon, Ghost, Slack, Discord, Brevo, and Telegram. One broadcast call reaches all of them.</p> <h2> Telegram approval flow </h2> <p>Before any post goes live, the bot sends a preview with Approve / Reject buttons. You review from your phone — one tap and it broadcasts. This is what makes it "assisted publishing" rather than "automated publishing."</p> <h2> Email onboarding </h2> <p>New signups get a 3-email sequence via Brevo, fired on first dashboard login. Plain text, signed with a human name. No behavioral conditionals — just time delays, which is all you need at launch.</p> <h2> How this post was published </h2> <p>Written in VS Code. Broadcast with one MCP tool call to <code>broadcast_post</code>. Social channels got a short version; dev.to, Hashnode, Ghost, and Brevo got this one.</p> <h2> What's next </h2> <ul> <li>Paddle billing (end-to-end test in progress)</li> <li>X and Threads adapters</li> <li>Scheduling support</li> </ul> <p>Try it: <a href="https://omnivocal.io" rel="noopener noreferrer">https://omnivocal.io</a></p> buildinpublic devjournal showdev socialmedia Omnivocal Dev Sun, 22 Mar 2026 21:34:15 +0000 https://forem.com/omnivocal/dev-log-1-building-in-public-with-mcp-jef https://forem.com/omnivocal/dev-log-1-building-in-public-with-mcp-jef <h2> What We're Building </h2> <p>OmniVocal is a tool that lets you publish blog posts to multiple platforms simultaneously via a single API call — or directly from your AI assistant using MCP (Model Context Protocol).</p> <p>This post is being published via the OmniVocal MCP server, running locally during development. The same flow works in production.</p> <h2> The Stack </h2> <ul> <li> <strong>Cloudflare Workers</strong> — edge API, zero cold starts</li> <li> <strong>D1</strong> — SQLite at the edge for posts, channels, API keys</li> <li> <strong>MCP server</strong> — exposes <code>broadcast_post</code> and <code>list_credentials</code> tools</li> <li> <strong>Adapters</strong> — per-platform publish logic (Hashnode, dev.to, Medium, GitHub, Bluesky)</li> </ul> <h2> Why MCP? </h2> <p>The interesting thing about building an MCP server for content distribution is that it collapses the workflow. Instead of:</p> <ol> <li>Open dashboard</li> <li>Paste content</li> <li>Select platforms</li> <li>Click publish</li> </ol> <p>You just describe what you want published in chat and the agent handles the rest — including formatting differences per platform.</p> <h2> What's Next </h2> <ul> <li>Paddle billing integration</li> <li>Medium and GitHub channels</li> <li>Scheduling support (Solo plan)</li> </ul> <p>More dev logs coming as we build.</p> cloudflare mcp buildinpublic devlog Omnivocal Dev Sun, 22 Mar 2026 21:34:03 +0000 https://forem.com/omnivocal/test-ua-iia https://forem.com/omnivocal/test-ua-iia <p>Hello UA.</p> Omnivocal Dev Sun, 22 Mar 2026 20:51:45 +0000 https://forem.com/omnivocal/hello-from-omnivocal-18h9 https://forem.com/omnivocal/hello-from-omnivocal-18h9 <p>Hello from OmniVocal — testing the setup and that everything works.</p> webdev buildinpublic devlog