MCP Is Mostly Bullshit

<a href="https:&#x2F;&#x2F;archive.ph&#x2F;nRsg3" rel="nofollow">https:&#x2F;&#x2F;archive.ph&#x2F;nRsg3</a>

Worth noting that Anthropic themselves are steering away from the &quot;MCP&quot; term - they&#x27;re using &quot;Claude custom integrations&quot;

sure there are hype around MCP, but how does that bridges to &quot;MCP Is Mostly Bullshit&quot;?<p>the title is more or less click bait, even author admits it<p>&gt; I am not even saying MCP is bad tech or useless. It’s just one way among others to provide context to AI assistants&#x2F;agents. If you have ever built an LLM-based application, you have more or less done something similar.<p>but beyond just similar, it&#x27;s very good to see there is a standard protocol that everyone adopts it, and the MCP will make the existing tools &amp; services immediate ready with better models being released in the future much easier

I think that MCPs are awesome.<p>I also think that they wont get much adoption, just like OSGi.<p>In general, people have shit for brains and MCP is just too abstract a concept for their tiny turds.

You can tell that MCP (model context protocol) was designed by those who haven&#x27;t thought about security or looked at a rigorous RFC (request for comment) document and instead done it for the &#x27;vibes&#x27;.<p>MCP has the equivalent security as the over-hyped and still error-prone JWT (Json Web Token) standard.<p>Both are horrific &#x27;standards&#x27; designed to cause lots of security incidents.