Chrome Devtools MCP now has an (experimental) CLI as well and can produce neat things like Lighthouse Audits.

https://github.com/ChromeDevTools/chrome-devtools-mcp/pull/1...

I've only used it a bit, but it's working well so far.

cool! needs to mature a bit, session sharing is a no-go for me as I need to run requests in parallel and it would interfere with each other.

The web is the only major platform that has a language monoculture to its detriment (i.e., not all problems are Javascript shaped). IMO the web ought to become multilingual (and become JS optional_ to further ensure its continued longevity and agility. Hopefully one day browser vendors will offer multiple runtime downloads (or something similar capability).

> i.e., not all problems are Javascript shaped

I’m having trouble coming up with a single Python-shaped problem that can’t be contained within JavaScript-shaped ecosystem.

WASM already offers this, for better or worse... There should be improved interop APIs for DOM access, but WASM is already very useful and even for directed UI control, "fast enough" a lot of the time. Dioxus, Yew and Leptos are already showing a lot of this to be good enough. That said, I would like to see a richer component ecosystem.

Embedded systems, consoles and mobile phones come to mind as well.

Even if you can go outside the blessed languages, it isn't without pain and scars.

All the embedded systems I've worked in have many languages you can use to compile whatever, burn, and run whatever you like. Consoles run game engines and programs written in all sorts of different languages. They don't care as long as they can execute the binary. Phones can run apps using many different languages (C, C++, Rust, Python, etc.).

Failed to read my second sentence.

There's a really nice typing plug-in for mypy that's been around a long time: https://github.com/typeddjango/django-stubs

It is very disappointing that these new type checkers don't support plug-ins, so things like django-stubs aren't possible. That means you're stuck with whatever is delivered with these new type checkers. It must be really difficult since none of them support plug-ins. Some of these newer type checkers promise support for Django, but you're stuck with what they (will) have on offer. Also, you'll likely want typing for other libs you might use.

Pyrefly's Django support is documented here: https://pyrefly.org/en/docs/django/

I believe Zuban also has some form of Django support, but I'm unable to locate the docs

The Agentic Lab is a nice community with people sharing their approaches and ideas similar to the OP.

https://www.skool.com/agentic

You're comparing apples, bananas, and pineapples while pretending they're all one thing. Switchblades are extremely effective (albeit expensive) anti-personnel (300 model) and anti-armor (600 model) drones. Shaheds are much larger, cheap, low on capabilities, but attritable used to attack fixed positions (e.g., buildings). These are all very different.

That's the point -- maybe the US should have bought apples instead of buying bananas.

I don't understand your point. Switchblades are (roughly) more akin to FPV (300 model) and Vampire drones (600 model) with reapect to size and payloads. Shahed style drones are roughly like like low end cruise missles. Different form factors and different capabilities. All of them are needed, but they're all very different.

A cruise missile is 3,000,000$ and a shahed drone is 50,000$ so if it’s even remotely the same capability it is an immense technological improvement over an expensive and slow to manufacture cruise missile.

You need a high/low capability that mixes all levels. For example, the Ukrainians and the Russians are both manufacturing very expensove cruise missles (Neptune/Iskander) and long range attack drones (shahed/fp-2/lute/etc). At any rate the original post I was responding to was comparing Switchblades to Shaheds, which is non-sensical.

What is the use case of a dumb, slow, suicide drone for the US army?

What’s the use case of a flying bomb that can be mass produced at little cost in days to weeks instead of months to years? Yeah tough to say really.

Well, what is it going to be used for?

Let’s compare it to Lancet then.

Lancets were miles better than Switchblades.

AFAIK Switchblades were used only in first months of the war then completely abandoned.

If only they had bought banana bombs!

> Then what's your excuse for why wasm, despite years of investment, is a dud on the web?

Depending on how you count, it took JS about 20 years and billions of dollars plowed into it to do the same, so why expect any less from wasm?

MCP eats lots of context (~20k tokens for chrome's). The more tokens you use needlessly, the faster your context rots (i.e., worse performance).

If you do that you'll reset browser state with whatever you're replacing. For example things like scroll position, if the user opened an accordion, etc. would all be lost.

> The correct way is to use EventSourcing, with each change to an otherwise immutable state being recorded as an Event, including a Delete (both of an Event and the Object).

Another great (and older) approach is adding temporal information do your traditional database, which gives immutability without the eventual consistency headaches that normally comes with event sourcing. Temporal SQL has their own set of challenges of course, but you get to keep 30+ years of relational DB tooling which is a boon. Event sourcing is great, but we shouldn't forget about other tools in our toolbelt as well!

I am using Temporal tables in SQL Server right now - I agree it's a bit of best of both worlds; but they are also painful to manage. I believe there could be a better solution without sacrificing SQL tools.

If you're on a Linux or Unix OS, a chroot jail might be a more lightweight solution. the chroot command essentially makes the chrooted directory look like the root dir. You need to set up all the directories claude can access (like /usr/bin or whatever). I haven't tried this yet, but I don't see any reason it wouldn't work. This solution would protect files outside your project from getting trashed, but not malicious data exfiltration.