101–110 of 200 posts
Not sure who the intended user is here? For frontend testing you actually do somewhat want the tests to break when the UI changes in major ways. And for other automation you'd be better off providing an actual API to use.
Could all of this be replaced simply by publishing OpenAPI (Swagger) spec and using universal swagger mcp client ??? This basically leaves up to the user to establish authenticated session manually. Assuming claude is smart enough to pick up API key from prompt/config, and can use swagger based api client, wouldnt that be the same?
That was everyone's first thought when MCP came out. Turns out it doesn't work too well since there is generally too many tools. People are doing interesting work in this space though
Earlier quoted context omitted.
But then, how are there so few commits in its profile graph? I suppose I may be admitting my ignorance of how public GitHub works, but still curious.
git require name and email to commit, but there's no cryptography involved there. Maybe GitHub won't green the lawn for unverified third party commits included in pushed commits?
Earlier quoted context omitted.
We should be focusing on llms using self discovery to figure out information. Can you expand? What does that mean, and why is the right (or better) path
Manually coding things is not how we get better AI. For AI to be truly useful in the area of figuring things out (i.e actually reasoning), one of the core components of a model would be building its own knowledge trees across multi modal information. So when you ask a model to do something, it should figure out how to do it on its own.
Two different goals.
I still don't understand MCP. If according to all the AI companies soon AI will replace devs than why bother with MCP?
I’ve haven’t used any MCP so far but as a disabled person I see use cases in accessibility for MCPs doing browser/smartphone automation. But any accessibility tool will be exploited by nefarious actors so I wonder how many main stream websites/apps would implement these MCP. Has anyone tried any MCP for improving accessibility?
How so?
> If I asked you to build a table and gave you a Home Depot you probably would have a harder time than if I gave you a saw, a hammer and some nails. I doubt that, first and not least because Home Depot stocks lumber.
Home Depot also sells tables.
Earlier quoted context omitted.
We mean if we could use this to instantiate an MCP for any website we're visiting.
Yea, I am planning a dev build of MCP-B which has access to user scripts apis. So technically you could `vibe code` and inject an MCP server into the target webpage
Or, in other words, it helps users get around all the bullshit that actually makes money for the business. Ads, upsells, cross-marketing opportunities. Engagement. LLMs help users avoid all that, and adding an MCP to your site makes it trivial for them.
Maybe I'm too cynical, but think about this: the technologies we needed for this level of automation became ubiquitous decades ago. This was, after all, also the original hype behind APIs, that burned bright on the web for a year or two - before everyone realized that letting people interact with webservices the way they want is bad for business, and everything closed down behind contracts backed by strong auth. Instead of User Agents connecting diverse resources of the web for the benefit of users, we got... Zapier. That's what the future of service-side MCPs on the web is.
But user scripts were always a way to let at least power users show a middle finger to "attention economy" and force some ergonomy out of web apps that desperately try to make users waste time. Giving users the ability to turn any website into MCP, regardless of whether that website wants it, will supercharge this. So for better or worse, that's where the future is.
Adversarial interoperability remains the name of the game.
I can see with my prophetic/logic eyes that free models will start to require captcha because of people start using MCP to automate browsers to use free LLMs. But captchas are ineffective against LLM so LLMs will fight automated LLMs from using them... Sounds like a very strange world of robots fighting robots
Earlier quoted context omitted.
But then, how are there so few commits in its profile graph? I suppose I may be admitting my ignorance of how public GitHub works, but still curious.
I was guessing Anthropic asked them to turn it off. Though why not also ask to delete the old activity of the person you bought the account from? Like: https://github.com/fotinakis/swagger-blocks/issues/3