Remix.run Logo
Remote MCP Support in Claude Code(anthropic.com)
118 points by surprisetalk 4 days ago | 55 comments
vessenes 8 hours ago | parent | next [-]

This is great. As someone who has made a few MCP servers in the last few months, MAN this spec moves fast; well ahead of Anthropic's internal support for it, and well ahead of documentation for implementation. It's like the Javascript community suddenly got automatic code creation agents, and went to town..

That said, the original spec needed some rapid iteration. With https support finally in relatively good shape, I hope we'll be able to take a year to let the API dust settle. Spec updates every three months are really tough, especially when not versioned, thoroughly documented, or archived properly.

zackify 9 minutes ago | parent | next [-]

It has been moving fast.

One weird thing I found a few weeks ago, when I added my remote MCP to Claude's integration tab on the website, I was getting OAuth errors.

Turns out they are requiring a special "claudeai" scope. Once I added that to my server, I was able to use it remotely in claude desktop!

I couldn't find any docs or reasons online for them requesting this scope.

Also, I have been using remote mcps in claude code for weeks with the awesome mcp-remote proxy tool. It's nice to not need that any longer!

Then as I'm writing a book currently on MCP Servers with OAuth, Elicitations come out! I'm rushing to update this book and be the best source for every part of the latest spec, as I can already see lots of gaps in documentation on all these things.

Huge shout out to VS Code for being the best MCP Client, they have support for Elicitations in Insiders already and it works great from my testing.

stingraycharles 7 hours ago | parent | prev | next [-]

Yeah, security really is an afterthought with most of these tools, but man the community is moving insanely fast — probably because most of these people are using these automation tools to develop their MCP servers in the first place.

It’s interesting to see other tools struggling to keep up. ChatGPT supposedly will get proper MCP client support “any day now”, but I don’t see codex supporting it any time soon.

Aider is very much struggling to adapt as well, as their whole workflow of editing and navigating files is easily replaced by MCP servers (probably better as well, as it provides much effective ways of reducing noise vs signal), so it’ll be interesting to see how tools adapt.

I’d love for Claude Code (or any tool for that matter) to fully embrace the agentic way of coding, e.g. have multiple agents specialize in different topics and some “main” agent directing them all. Those workflows seem to be working really well.

ljm 2 hours ago | parent | next [-]

The real security issue is around the use of ‘YOLO mode’ where you just let the agent invoke tools in a completely unattended manner. It’s not much different than slapping sudo in front of every shell command or running as root.

People are going to continue doing that because these agentic tasks can take some time to run and checking in to approve a command so often becomes an annoyance.

I can’t see a way around that except to have some kind of sandboxing or a concept of untrusted or tainted input rather than treating all tokens as the same. Maybe a way of detecting if the response of a tool is within a threshold of acceptability within the definition of the MCP (which is easier with structured output), which is used to force a manual confirmation or straight up rejection if it’s deemed to be unusual or unsafe.

alvis 15 minutes ago | parent [-]

It's indeed an issue. I love codex that it contains everything in a sandbox and I can review what has changed. It's proper and I've much better idea what's going on.

That said, I ditched codex for claude code... Sorry open ai. No MCP and no way to interact during execution is a huge drawback.

vessenes an hour ago | parent | prev | next [-]

The aider slowdown is a real bummer. I’d love to have Claude code UI with the model choice aider gets me, but I’m not willing to give up tool integration.

Maxious 6 hours ago | parent | prev [-]

https://github.com/ruvnet/claude-code-flow adds some of the multiagent features ontop

stingraycharles 5 hours ago | parent [-]

Yeah that’s what I’m experimenting with, but I think it’s overengineered, especially with the whole dogmatic SPARC approach. I’m personally a more minimalistic person, and I would prefer it to be natively integrated into the app and being able to define exactly the (system) prompts for each of the agents.

jmsdnns an hour ago | parent | prev | next [-]

> Spec updates every three months are really tough, especially when not versioned, thoroughly documented, or archived properly.

Couldnt AI help with that..?

vessenes an hour ago | parent [-]

Ironically the updates make it hard to get llm coding support for it. Very ironically.

CharlieDigital 5 hours ago | parent | prev | next [-]

    > Javascript community suddenly got automatic code creation agents, and went to town.
I've been working on an MCP server[0] that let's LLMs safely and securely generate and execute JavaScript in a sandbox including using `fetch` to make API calls. It includes a built in secrets manager to prevent exposing secrets to the LLM.

I think this unlocks a lot of use cases that require code execution without compromising security. Biggest one is that you can now ask the LLM to make API calls securely because the JS is run in a C# interpreter with constraints for memory, time, and statement limits with hidden secrets (e.g. API keys).

The implementation is open source with sample client code in JS using Vercel AI SDK with a demo UI as well.

[0] https://github.com/CharlieDigital/runjs

rvnx 6 hours ago | parent | prev [-]

The crazy thing about things moving fast is that people bought Cursor for hundreds of millions when it is already outdated by Claude Code. Very foolish by the purchasers but very smart for the founders

diggan 4 hours ago | parent [-]

I think this is why we're seeing founders selling so quickly with these startups. You could wait some weeks or months to sell higher, but seems chances are higher that whatever you've built is outdated by then so why risk it?

schappim 7 hours ago | parent | prev | next [-]

This is great news; remote MCP support should be open and accessible.

For what it’s worth, I’ve been using WitsyAi: it’s fully free, open source, and serves as a universal desktop chat-client (with remote MCP calling). You just need to BYO API keys.

Remote MCPs are close to my heart; I’ve been building a “Heroku for remote MCP tools” over at Ninja[2] to make it easy for people to spin up and share MCP tools without the usual setup headaches.

Lately, I’ve also been helping folks get started with MCP development on Raspberry Pi. If you’re keen to dive in, feel free to reach out [3].

[1] https://witsyai.com

[2] https://ninja.ai

[3] https://calendly.com/schappi/30min

aberoham 7 hours ago | parent [-]

I've found this from an old-school systems geek to be useful https://github.com/giantswarm/mcp-debug -- especially its REPL mode

mark_undoio 3 hours ago | parent [-]

Thanks for this - I've been using the MCP Inspector https://modelcontextprotocol.io/docs/tools/inspector but find it doesn't really fit my workflow.

I like the fact this mcp-debug tool can present a REPL and act as a mcp server itself.

We've been developing our MCP servers by first testing the principle with the "meat robot" approach - we tell the LLM (sometimes just through the stock web interface, no coding agent) what we're able to provide and just give it what it asks for - when we find a "tool" that works well we automate it.

This feels like it's an easier way of trying that process - we're finding it's very important to build an MCP interface that works with what LLMs "want" to do. Without impedance matching it can be difficult to get the overall outcome you want (I suspect this is worse if there's not much training data out there that resembles your problem).

joshwarwick15 9 hours ago | parent | prev | next [-]

List of servers to connect to here: https://github.com/jaw9c/awesome-remote-mcp-servers

movedx01 6 hours ago | parent | next [-]

That's great! It would be even better if one of the features included in the table was whether given MCP supports OAuth Dynamic Client Registration, which optional in the MCP standard.

hirsin 3 hours ago | parent [-]

The MCP server technically doesn't support DCR. The authorization server for the MCP server does, which is a minor distinction.

Have you seen significant need for this? I've been trying to find data on things like "how many MCP clients are there really" - if it takes off where everything is going to be an MCP client && dynamically discovering what tools it needs beyond what it was originally set up for, sure.

digitcatphd 9 hours ago | parent | prev [-]

Thanks!!!

indigodaddy an hour ago | parent | prev | next [-]

If you’re on the Claude Pro plan and on a forced break, this tool might be useful:

https://github.com/robertpiosik/CodeWebChat

rane 3 hours ago | parent | prev | next [-]

How is this different from the existing MCP support in Claude Code?

CharlesW an hour ago | parent | prev | next [-]

Which MCP servers are Claude Code must-adds for coding (vs. ancillary tasks like managing Jira)? For example, @kazuph/mcp-fetch has been useful.

altvsnda an hour ago | parent | next [-]

Context7 is quite useful for coding, imo

CharlesW 12 minutes ago | parent [-]

Looks extremely helpful, thank you!

jasonjmcghee an hour ago | parent | prev [-]

I searched for this mcp server just to see what it was, and afaict Claude Code has all of these features built in already.

What is the additional functionality?

CharlesW an hour ago | parent [-]

You may be right, I'm fairly new to Claude Code. CC itself summarized the difference this way: "Use MCP Fetch [the tool I refernced] when you need complete content or images, use WebFetch [CC's built in tool] when you want AI to extract specific information from a page."

jasonjmcghee an hour ago | parent [-]

Ah- I missed it extracts images from the pages. I don't think Claude does this. But you can fetch pages and give it images (separately).

wewewedxfgdf 2 hours ago | parent | prev | next [-]

You cannot unsee what you see when the Claude logo animates.

4oo4 2 hours ago | parent [-]

I feel like they had to have read this and added that out of spite.

https://velvetshark.com/ai-company-logos-that-look-like-butt...

dr_kretyn an hour ago | parent [-]

THANK YOU! I really thought I have a perversive mind looking at them all. At least, I'm not the only one. (Anthropic is definitely the butt of them all.)

WhitneyLand 3 hours ago | parent | prev | next [-]

On a related note what’s the hesitation for Anthropic/OpenAI to open up their mobile apps to diy mcp servers?

conception 3 hours ago | parent [-]

Probably app approvers and dealing with that muck?

anotherpaul 4 hours ago | parent | prev | next [-]

If I understand correctly MCP is the interface to connect any API to LLMs.

Now a lot of people use it to add context to their model. And also tool calls?

I am using continue.dev not Claude but I imagine this tech stack will be ported everywhere.

As a python Dev I don't quite yet understand though how and what service I should be running. Or be using. Tbh. Can anyone ELI5?

jasonjmcghee an hour ago | parent | next [-]

You can give LLMs useful tools to have more rich interactions with them - help them help you more.

They're also fun and easy to build.

Here's one I made - it wraps the vscode debugger: https://github.com/jasonjmcghee/claude-debugs-for-you

I've specifically tested it with continue.dev so it might serve as a useful example / template.

beatthatflight 3 hours ago | parent | prev [-]

easiest one to get going with is to add the Playwright MCP. As a python dev you might have used it to do test automation? Anyway, it gives your tool eg Cursor, Claude Code access to the browser and automation using playwright. Meaning it can literally load a page to confirm its own change just had the desired effect.

The blender one is also fun as a starting point, if you do any 3d modelling (or even if you don't).

nvahalik 4 hours ago | parent | prev | next [-]

Used this Friday to have Claude do some stuff but telling it to read a Linear ticket and make appropriate changes. Not perfect but saved me 15 minutes.

max_on_hn 32 minutes ago | parent [-]

If you like that workflow you might love the tool[0] which I built specifically to support it: CheepCode connects to Linear and works on tickets as they roll in, submitting PRs to GitHub.

[0] https://cheepcode.com

Aeolun 5 hours ago | parent | prev | next [-]

Unless you try to add Github, and it fails everywhere :/

iddan 9 hours ago | parent | prev | next [-]

I'm using https://xander.bot, which already leverages MCP connection between Claude Code and Linear. Really awesome stuff, how it knows how to gather context from Linear.

michaericalribo 5 hours ago | parent | prev | next [-]

When will Gemini release a desktop app and enable MCP and coding agents??

fizx 13 minutes ago | parent | next [-]

Cursor + Gemini is rather good.

mark_undoio 3 hours ago | parent | prev | next [-]

I assume you're aware but there are some open source assistants that can get you this (at least to some extent).

E.g. Codename Goose https://block.github.io/goose/docs/quickstart/

Which I think supports Gemini among with all the other major AI providers, plus MCP. I have heard anecdotally that it doesn't work as well as Claude Code, so maybe there are additional smarts required to make a top notch agent.

I've also heard that Claude is just the best LLM at tool use at the moment, so YMMV with other models.

artdigital 5 hours ago | parent | prev [-]

Not a desktop app, but Gemini has “Jules” which is their autonomous coding agent

You hook up your GitHub repo and it’ll clone it, setup the env and then work on the task you give it

richardblackbox 8 hours ago | parent | prev | next [-]

This is great, we have recently released on the blackboxai vscode agent a live call with the coding agent

Will be trying the MCP with the live call as well, i think it should work

geeunits 5 hours ago | parent | prev [-]

I put 10 months of work into MCP. Been a Max 20 x subscriber since Day 1. Spent thousands via the API previously. Anthropic randomly banned my account with no warning or explanation last week. I've had to spent significant time and resources moving AWAY from MCP.

Do not get comfy with any of these protocols or companies. They can destroy your workflows and livelihood in an instant.

jasonjmcghee an hour ago | parent | next [-]

Reach out to support. Unless you were doing something egregiously against terms of use, I doubt you'll have trouble getting your access restored.

Also- Claude desktop is honestly one of the more minimal implementations of an mcp client with a bunch of missing features. Just search for mcp clients. There are many you can use.

geeunits 7 minutes ago | parent [-]

I did. No response. Filled out the form, no response. I've genuinely no idea what I could've done, outside of complain about the palantir xs scripting issue. Quite disheartening. Thanks for the tip!

Aeolun 5 hours ago | parent | prev [-]

What does MCP has anything to do with Anthropic or Claude Max? I mean, they made the protocol, but it works everywhere.

geeunits 5 hours ago | parent [-]

They invented the protocol? I used Claude Desktop extensively in testing. No other local clients currently interop very well with MCP Tooling. They're more of an afterthought for other clients.

Aeolun 5 hours ago | parent [-]

Sure, but it’s not suddenly useless to you because Anthropic restricts you. Those 10 months of building MCP servers are still of use, even if I can only imagine how frustrating losing Claude Max must be (recent x20 subscriber here)

geeunits 4 hours ago | parent [-]

The difference is, they were being used. Now they'll only be of use further down the line, and I have to readjust my daily workflow. Just because. It's highly frustrating and makes me regret putting so much effort in embracing 'the Anthropic way' when a rug pull can come out of nowhere. I wish I'd put that effort into my own tooling instead. Just a fair warning to anyone getting comfortable.

Aeolun 4 hours ago | parent | next [-]

I guess I’m just riding the wave while I can. I’d be pissed if the rug is pulled, but can’t do much more than move on.

owebmaster 3 hours ago | parent | prev [-]

What were you doing that got you banned?

geeunits 3 hours ago | parent [-]

I have no idea. I haven't altered my workflow or projects.