<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: sloth</title>
    <description>The latest articles on DEV Community by sloth (@sloth-os).</description>
    <link>https://dev.to/sloth-os</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4077403%2F380abc46-8e95-41cc-8362-6ef8afd3e9d1.png</url>
      <title>DEV Community: sloth</title>
      <link>https://dev.to/sloth-os</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/sloth-os"/>
    <language>en</language>
    <item>
      <title>Build One Bot, Reach Every Chat: A Universal WebSocket Bridge for IM Channels</title>
      <dc:creator>sloth</dc:creator>
      <pubDate>Wed, 19 Aug 2026 03:47:57 +0000</pubDate>
      <link>https://dev.to/sloth-os/build-one-bot-reach-every-chat-a-universal-websocket-bridge-for-im-channels-4jpm</link>
      <guid>https://dev.to/sloth-os/build-one-bot-reach-every-chat-a-universal-websocket-bridge-for-im-channels-4jpm</guid>
      <description>&lt;p&gt;Every chat platform reinvents the wheel. WeChat has its own API, Lark has another, Telegram another still — and if you're building a bot, an agent, or any service that needs to talk to people across IM channels, you end up writing and maintaining a separate integration for each one. The plumbing differs, the auth flows differ, the message formats differ, and every new channel means another chunk of bespoke code in your app.&lt;/p&gt;

&lt;p&gt;I got tired of that, so I built OpenClaw Bridge — a universal WebSocket bridge server that sits between your client and any number of IM backends. The idea is simple: your app speaks one protocol over a single WebSocket, and the bridge translates that into whatever each backend channel expects. Here's the shape of it:&lt;/p&gt;

&lt;p&gt;Client ──[WebSocket]──▶ Bridge Server ──[ChannelPlugin API]──▶ Backend Channels&lt;/p&gt;

&lt;p&gt;The key design decision is that the bridge contains zero plugin-specific code. It doesn't know how WeChat logs in, how Lark formats media, or how Telegram handles reply threading. Instead, it works with any channel that implements a standard ChannelPlugin interface. Install a plugin, drop its credentials into config.json, and it just works — WeChat, Lark, Telegram, Mattermost, IRC, QQ, WhatsApp, and whatever comes next.&lt;/p&gt;

&lt;p&gt;In this post I'll walk through why a bridge like this is worth building, how the envelope protocol keeps clients channel-agnostic, and how a subscription model lets multiple clients share one set of connected accounts without tripping over each other. If you've ever copy-pasted your fifth nearly-identical IM integration, this one's for you.&lt;/p&gt;

&lt;p&gt;For more information, visit &lt;a href="https://github.com/sloth-os/openclaw-bridge" rel="noopener noreferrer"&gt;https://github.com/sloth-os/openclaw-bridge&lt;/a&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>productivity</category>
      <category>opensource</category>
      <category>node</category>
    </item>
    <item>
      <title>mm-gateway: One Provider-Neutral API for Image, Video, and Music Generation</title>
      <dc:creator>sloth</dc:creator>
      <pubDate>Fri, 14 Aug 2026 09:32:30 +0000</pubDate>
      <link>https://dev.to/sloth-os/mm-gateway-one-provider-neutral-api-for-image-video-and-music-generation-2k5h</link>
      <guid>https://dev.to/sloth-os/mm-gateway-one-provider-neutral-api-for-image-video-and-music-generation-2k5h</guid>
      <description>&lt;p&gt;Every generative-AI app I worked on hit the same wall. Pick a backend — OpenAI for images, Volcengine for video, Mureka for music — and before long the app is &lt;em&gt;full&lt;/em&gt; of that provider's SDK quirks: its field names, its sync-vs-async polling loop, its error shapes. Then if want to swap one backend, or add a second one for failover, and it's a rewrite.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://github.com/sloth-os/mm-gateway" rel="noopener noreferrer"&gt;mm-gateway&lt;/a&gt;&lt;/strong&gt; is an open-source Python gateway that sits in front of that mess. One provider-neutral contract — over &lt;strong&gt;13 backends&lt;/strong&gt;:&lt;br&gt;
   OpenAI · Google · xAI · DashScope · Volcengine · Flux · Stability · ElevenLabs · MiniMax · Mureka · ACE-Step · OpenRouter · UdioAPI&lt;/p&gt;

&lt;p&gt;The idea is simple: &lt;strong&gt;provider wire formats never appear in application code.&lt;/strong&gt; Every request goes through a strict, modality-specific envelope — an ordered list of &lt;em&gt;typed input parts&lt;/em&gt; plus &lt;em&gt;provider-neutral parameters&lt;/em&gt; — and each backend adapter translates that to its native SDK or REST shape.&lt;/p&gt;

&lt;p&gt;For more information, visit &lt;a href="https://github.com/sloth-os/mm-gateway" rel="noopener noreferrer"&gt;https://github.com/sloth-os/mm-gateway&lt;/a&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>python</category>
      <category>opensource</category>
      <category>programming</category>
    </item>
  </channel>
</rss>
