jngiam1

What if these models were told to clean up their tracks?

I totally agree, we’ve been working with enterprises and MCP is the defacto way they are using agents with data.

We build a gateway (MintMCP) and I’d love to connect and exchange notes.

We’re excited about XAA, it will simplify many flows

I don't get why everything is not marked as sensitive in env vars by default instead.

MCP tools. We're doing some MCP bundling and giving it here, pretty cool stuff.

This is powerful. Combined with MCPs, you can pretty much automate a ton of work.

We built a very similar thing! Also with git, very nice- if you’re looking for an enterprise ready version of this, hit me up

Love to discuss and see how we can make this more standard

Honestly, easier with MCP straight up: https://gmail.mintmcp.com/ https://gcal.mintmcp.com/ https://gdocs.mintmcp.com/ https://gsheets.mintmcp.com/

(all pass through)

Exactly. and even if so, how are you going to safe guard tool access?

Imagine your favorite email provider has a CLI for reading and sending email - you're cool with the agent reading, but not sending. What are you going to do? Make 2 API keys? Make N API keys for each possible tool configuration you care about?

MCPs make this problem simple and easy to solve. CLIs don't.

I don't think OpenClaw will last that long without security solved well - and MCPs seem to be obvious solution, but actively rejected by that community.

I think if you want background agents with sandboxes and well scoped permissions, you want MCP to be your data protocol and security layer.

If you’re vibing and doing the open claw thing without any security concerns; then you’re absolutely right.

The Skills I have for Claude are all based on personal preferences and reflects the setup I have going. It's a way to narrow the probability space to the specific set which works really well for me.

i've a simple setup with Claude Code and MCPs; and i get real benefits from better task mgmt, email mgmt, calendar, health/food/fitness tracking, working together with claude on tasks (that go into md files).

i don't think we need ClawdBot, but we do need a way to easily interact with the model such that it can create long term memories (likely as files).

The mirror test is cool!

Store/ version with git, throw Claude code at it, and it’ll be amazing

Lutra is a native code-mode AI agent that: (a) converts all tools into functions available in a coding environment, (b) uses LLMs to produce code that orchestrates across those tools, (c) has a custom stateless code interpreter to make it all work securely and well.

There’s too much rage baiting on the internet now; the headlines that take the extreme position get reshared, while the truth is more in the middle.

This is the key. MCP encapsulates tools, auth, instructions.

We always need something for that - and it needs to work for non tech users too

I hypothesize that these AI agents are all likely higher than human performance now.

MCP gateway: https://mintmcp.com

There's an agent monitor which intercepts requests either using a LLM proxy or hooks, that gives you full telemetry into the agents + MCPs used. And a MCP gateway that enables centralized deployment and securing of MCP.

I've been hoping that Claude Code on the Web also works with MCPs; so I can start getting it to do things beyond just coding. It's pretty awesome to use Git as a source of memory/tracking what's going on and pull requests as a way to build in a human-in-the-loop review flow.

We do have self-hosted options, and are still thinking through OSS strategies. If you've suggestions on that angle, would love to learn / read more.

+1 many tools can be run stateless / have state in a external db, and we have ways to support running that efficiently as well.

It's really messy now. ChatGPT in particular makes it really hard; it turns out that Custom GPTs with actions can do pretty well, if you bridge MCP tools into actions; but setting any of these up is a pita.

The LLMs are also really bad at generating correct code for OAuth logic - there are too many conditions there, and the DCR dance is fairly complicated to get right.

Shameless plug: we're building a MCP gateway that takes in any MCP server and we do the heavy lifting to make it compatible with any client on the other end (Claude, ChatGPT - even with custom actions); as a nice bonus it gives you SSO/logs as well. https://mintmcp.com if you're interested.

I got so used to having Claude Code read some of my MCP tools, and was bummed to see that it couldn't connect to them yet on the web.

Pretty cool though! Will need to use it for some more isolated work/code edits. Claude Code is now my workhorse for a ton of stuff including non-coding work (esp. with the right MCPs)

Strong agree here

MCPs have a larger impact beyond the terminal - you can use it with ChatGPT, Claude Web, n8n, LibreChat, and it comes with considerations for auth, resources, and now even UI (e.g., apps-sdk from OpenAI is on MCP).

If we're considering primarily coding workflows and CLI-based agents like Claude Code, I think it's true that CLI tools can provide a ton of value. But once we go beyond that to other roles - e.g., CRM work, sales, support, operations, finance; MCP-based tools are going to have a better form factor.

I think Skills go hand-in-hand with MCPs, it's not a competition between the two and they have different purposes.

I am interested though, when the python code in Skills can call MCPs directly via the interpreter... that is the big unlock (something we have tried and found to work really well).

Made a marketplace for a secret scanner hook we built: https://github.com/mintmcp/agent-security

I have been thinking that the appropriate solution here is to detect when one of the legs is appearing to be a risk and then cutting it off if so.

You don’t want to have a blanket policy since that makes it no longer useful, but you want to know when something bad is happening.

I also ask the agent to keep track of what we're working on in a another md file which it save/loads between clears.

Curious to learn more- have a pointer to it?