WEBVTT 00:00:00.020 --> 00:00:04.780 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:05.200 --> 00:00:10.020 This is episode 491, recorded on Tuesday, August 11th, 2026. 00:00:10.520 --> 00:00:11.540 I'm Calvin Hendryx-Parker. 00:00:11.800 --> 00:00:12.560 And I'm Michael Kennedy. 00:00:13.000 --> 00:00:14.300 Hey, Michael, it's great to see you. 00:00:14.620 --> 00:00:16.240 It's great to be back. I missed you, Calvin. 00:00:16.430 --> 00:00:21.180 I missed you too. A week off, a week with no Michael was a little, you know, somehow I got through. 00:00:22.580 --> 00:00:24.640 Let's get this thing started. We've got a lot to talk about. 00:00:24.710 --> 00:00:28.160 I see some cool stories on the docket for today. 00:00:28.560 --> 00:00:30.940 But first, this episode is sponsored by XWeather. 00:00:31.340 --> 00:00:38.340 XWeather combines enterprise-grade weather intelligence with agent-ready APIs, natural language capabilities, and an MCP server. 00:00:38.680 --> 00:00:44.540 So your agents can adopt workflow, automate responses, and make better decisions based on real-world conditions. 00:00:44.900 --> 00:00:51.220 Michael's going to tell you more later in the show, but get started for free at pythonbytes.fm/xweather. 00:00:51.620 --> 00:00:57.480 If you want to connect with Michael or I on the socials, we are on Mastodon, BlueSkyX, and LinkedIn. 00:00:57.520 --> 00:01:01.100 You can also follow the show at MassathonBlueSkyNX. 00:01:01.720 --> 00:01:04.900 Join us on YouTube for the live session of this. 00:01:04.980 --> 00:01:08.360 So at pythonbytes.fm/live to be part of the live audience. 00:01:08.880 --> 00:01:12.660 We usually record this on Tuesdays at 7 a.m. Pacific, 10 a.m. Eastern. 00:01:13.080 --> 00:01:14.820 Older video versions are there too. 00:01:15.210 --> 00:01:18.920 We record without a net so you can watch us live and join in all the fun. 00:01:19.280 --> 00:01:27.040 Finally, if you want an artisanal digest of every week's shows and the notes from the email form, add your name and email to our friends of the show list. 00:01:27.120 --> 00:01:29.560 We promise we will never share it with anyone else. 00:01:30.000 --> 00:01:35.720 You're just going to get cool messages from Michael and myself about all the show notes here, which are pretty good. 00:01:35.860 --> 00:01:37.860 You could almost read the show notes and not listen to us. 00:01:38.000 --> 00:01:40.340 But I know everyone loves hearing our voices. 00:01:40.710 --> 00:01:42.540 And speaking of, Michael, let's kick it off. 00:01:42.590 --> 00:01:43.320 What do you got for us first? 00:01:43.660 --> 00:01:44.900 Let me ask you a question. 00:01:44.950 --> 00:01:47.300 I know you do a fair amount of AI stuff. 00:01:47.680 --> 00:01:48.000 A little bit. 00:01:48.520 --> 00:01:49.700 Yeah, a little bit. 00:01:49.910 --> 00:01:50.380 And me too. 00:01:50.780 --> 00:01:51.140 I dabble. 00:01:52.280 --> 00:01:54.280 It's probably sad, but I'm going to stick with it for a week or two. 00:01:55.050 --> 00:01:55.500 Give it a go. 00:01:55.800 --> 00:01:59.960 This one, you have to ask yourself, are you ready to be judged? 00:02:00.480 --> 00:02:02.500 And sort of what's your confidence level? 00:02:02.890 --> 00:02:07.340 So what I'm talking about is something called Claude Code Insights. 00:02:07.620 --> 00:02:09.200 Have you experimented with this, Calvin? 00:02:09.440 --> 00:02:11.760 I've not run the slash insights command yet. 00:02:11.940 --> 00:02:14.680 I've been totally in the world of AgentsView from last time. 00:02:14.980 --> 00:02:15.260 I know. 00:02:15.380 --> 00:02:16.440 I've been all about AgentsView. 00:02:16.510 --> 00:02:16.660 Same. 00:02:16.660 --> 00:02:20.040 And I'm so bugged that I can't get, I know I can. 00:02:20.100 --> 00:02:28.960 And I have not yet set up AgentsView to unify my Mac laptop and desktop because I'm only seeing half of my view. 00:02:29.280 --> 00:02:29.460 Yeah. 00:02:29.780 --> 00:02:30.080 Yeah. 00:02:30.320 --> 00:02:31.620 So I'm going to set that up at some point. 00:02:31.710 --> 00:02:32.260 But you know what? 00:02:32.560 --> 00:02:37.140 There's a few things more important than that, but not that many things I'm like, I want to do this. 00:02:37.220 --> 00:02:37.620 Let's check it out. 00:02:37.920 --> 00:02:38.700 So Claude Code Insights. 00:02:38.840 --> 00:02:39.200 What is this? 00:02:39.520 --> 00:02:44.200 So Claude Code Insights is you just type slash insights, press enter, nothing else. 00:02:44.380 --> 00:02:48.320 Now, annoyingly, this does not work in the VS Code Variants plugin. 00:02:49.000 --> 00:02:52.340 You can say slash compact, clear, slash usage. 00:02:52.800 --> 00:02:56.720 All those work in the VS Code version, but this has to literally be in the terminal version. 00:02:57.020 --> 00:02:58.740 Does it have anything to do with the terminal? 00:02:59.040 --> 00:03:00.060 No, absolutely not. 00:03:00.400 --> 00:03:01.400 But it still has to be there. 00:03:01.560 --> 00:03:01.900 Don't know why. 00:03:02.420 --> 00:03:03.520 But it's fine. 00:03:03.700 --> 00:03:10.520 So if you're not using the terminal version of Claude Code, you have to temporarily go down and use that to get this to work, okay? 00:03:10.800 --> 00:03:13.080 So what you do, like I said, you just type slash insights. 00:03:13.260 --> 00:03:19.880 And what it does is it will look through all of your sessions, which are stored on your local machine, not in the cloud, just on your machine. 00:03:20.020 --> 00:03:24.340 So you've got to do it on a machine that you use Claude Code a lot to get anything meaningful from it. 00:03:24.600 --> 00:03:26.240 And it will assess you, Calvin. 00:03:26.640 --> 00:03:29.540 It'll tell you how you do rather than how it's done. 00:03:30.200 --> 00:03:32.000 Am I guiding you well enough? 00:03:32.260 --> 00:03:32.940 So what I've done. 00:03:32.940 --> 00:03:33.480 Am I treating you nicely? 00:03:34.140 --> 00:03:34.500 Exactly. 00:03:34.920 --> 00:03:40.600 What I've done is I'm like, all right, well, if I'm going to talk about this, let me just share one of my Claude Code insights. 00:03:40.620 --> 00:03:43.380 So I ran that and I put it up on the internet. 00:03:43.780 --> 00:03:44.000 Wow. 00:03:44.580 --> 00:03:45.460 That's being very vulnerable. 00:03:45.900 --> 00:03:46.300 I know. 00:03:46.460 --> 00:03:48.200 See, I'm open here. 00:03:48.380 --> 00:03:49.680 I may regret it, but I'm open here. 00:03:51.120 --> 00:03:51.780 Here's the thing. 00:03:51.840 --> 00:03:55.120 I believe this is a 30-day window, even though it says the time is bigger. 00:03:55.820 --> 00:03:57.440 But the docs say it's 30 days. 00:03:57.580 --> 00:03:59.080 I don't know what the actual time frame is. 00:03:59.080 --> 00:04:00.220 It doesn't really matter that much. 00:04:00.500 --> 00:04:04.480 It talks about how many sessions there were, how many messages, all those kind of things. 00:04:04.580 --> 00:04:18.739 But then it gives you an overall assessment, what went well, what went bad, and what's super interesting and why you might actually care rather than just morbid curiosity of what AI thinks of you is it gives you a bunch of things to make it work better. 00:04:18.739 --> 00:04:22.540 It says, look, when we tried these things, systematically, this is not working that well. 00:04:22.830 --> 00:04:25.740 So here, paste this into Claude Code to create a skill that makes it better. 00:04:26.220 --> 00:04:26.820 That's cool. 00:04:26.880 --> 00:04:34.620 Yeah, so it's kind of like a meta thing and sort of a counseling because it also gives a little bit of insight into itself. 00:04:35.110 --> 00:04:43.540 Now, you may not want to put your Claude Code insight on the internet because it has detailed information about your projects. 00:04:44.060 --> 00:04:53.280 It could say, oh, yeah, I worked at big tech company X and you were really successful in helping track down that huge vulnerability or whatever. 00:04:53.930 --> 00:04:55.260 Then describe it, you know what I mean? 00:04:55.460 --> 00:05:09.880 Yeah, that's no good. That's a reason I can't publish any of my agent's views because it definitely tells you very detailed information about what's going on in there. I like that it's suggesting to you basically you should use superpowers in your plan, phase, verify, commit. 00:05:10.680 --> 00:05:24.720 Yes, exactly. It does tell me that. So here, I'll read you a few fragments of it so you get a sense. And then you all can go run slash insights either as you're listening right now or later. So what it does is it just generates an HTML file. Like that's why I was complaining it doesn't work in the VS Code variants. 00:05:24.960 --> 00:05:30.240 or any other UIs because literally all it goes is thinking, here's your HTML file, please open it up. 00:05:30.520 --> 00:05:31.660 It's like, that can work anywhere. 00:05:31.740 --> 00:05:32.660 I don't know what you're talking about. 00:05:32.660 --> 00:05:34.240 But anyway, it gives you an at a glance. 00:05:34.320 --> 00:05:35.080 Like what's working? 00:05:35.420 --> 00:05:37.760 You work in phases rather than one-shot requests. 00:05:38.180 --> 00:05:42.980 Number plan files, one phase at a time with verification gates between them. 00:05:43.280 --> 00:05:43.720 That's discipline. 00:05:44.640 --> 00:05:50.640 That discipline is what let you drive ty diagnostics from 363 to zero across 25 commits without destabilizing. 00:05:50.820 --> 00:05:54.140 A 1,399 test suite. 00:05:54.560 --> 00:06:02.080 And why your Sentry work closes the full loop from raw ID issue to root cause, fix, test, change log, and resolve issue. 00:06:02.520 --> 00:06:09.200 By the way, I've been using the Sentry MCP, Sentry past sponsor, possible future sponsor, definitely sponsor. 00:06:09.560 --> 00:06:10.260 General good folks. 00:06:10.660 --> 00:06:11.460 Really like Sentry. 00:06:11.640 --> 00:06:11.760 Yeah. 00:06:12.020 --> 00:06:18.280 And so I've hooked it up to report errors when something goes wrong, and it will just log those with details. 00:06:18.590 --> 00:06:21.780 And so now you can just go, hey, Claude, there's an issue over at Sentry. 00:06:22.200 --> 00:06:22.760 Look into it. 00:06:23.020 --> 00:06:23.520 Propose a fix. 00:06:24.000 --> 00:06:25.840 walk away and it's really, really cool. 00:06:26.120 --> 00:06:35.640 So that's what it's talking about, right? So that's what's working. And it said, well, what's going on? Well, on Claude's side, the biggest cost was asserting correctness instead of verifying, like saying, yes, that's how we're saying. No, it doesn't work that way, Claude. 00:06:36.280 --> 00:06:40.220 Stop saying that. Go check. Oh, you're absolutely right. You know, remember, that's not a code review, is it? 00:06:40.440 --> 00:06:41.440 It's so self-aware. 00:06:42.280 --> 00:06:50.000 It is self-aware. And the quick ones to try, turn your plan phase, verify, commit loop into a custom skill. This is where Calvin is suggesting that I just adopt 00:06:50.440 --> 00:06:51.600 superpowers. You should just give it. 00:06:52.020 --> 00:07:00.600 I have used superpowers and I like really well how it came out, but I feel less connected to the code when I'm using superpowers. 00:07:01.020 --> 00:07:03.520 In that it just goes, I got this. 00:07:04.000 --> 00:07:04.260 See you later. 00:07:04.560 --> 00:07:07.500 And then like two hours later, here's your output. 00:07:08.600 --> 00:07:09.760 Where's the flow that I have? 00:07:09.820 --> 00:07:14.580 It checks in with me just a little bit or at least it narrates while it's going. 00:07:14.640 --> 00:07:15.320 I can hold on. 00:07:15.720 --> 00:07:16.680 Why did you just say that? 00:07:16.680 --> 00:07:17.100 You know what I mean? 00:07:17.760 --> 00:07:18.080 I don't know. 00:07:18.600 --> 00:07:20.200 Maybe I will someday give in to superpowers. 00:07:20.440 --> 00:07:22.460 But I have a very similar flow to it nonetheless. 00:07:23.340 --> 00:07:24.080 Ambitious workflow. 00:07:24.400 --> 00:07:26.380 Your sentry to triage loop was deterministic. 00:07:26.580 --> 00:07:31.440 I did this whole mega thing where I went through the Talk Python mobile apps. 00:07:31.880 --> 00:07:35.680 They had years of errors that I was like, I cannot deal with these. 00:07:35.680 --> 00:07:41.920 And I cannot do – this is too subtle for me to fix without making it worse in Flutter and then getting it back in the App Store. 00:07:42.060 --> 00:07:45.920 And I eventually just broke down and said, okay, I'm going to – Claude and I are figuring this out. 00:07:46.380 --> 00:07:49.420 That's a perfect kind of tedium for these models. 00:07:50.060 --> 00:07:50.900 Yeah, it was so good. 00:07:50.960 --> 00:07:51.480 It was so good. 00:07:51.520 --> 00:07:55.260 It will even open up the simulator and automate it to track down the issues and so on. 00:07:55.540 --> 00:08:03.940 So there's a lot more than I'm reading here, but it tells you what you worked on, how you used Claude Code, impressive things, what went wrong on the horizon, all these things. 00:08:04.120 --> 00:08:09.720 So it tells you here's all your projects, and this is certainly where you want to be a little less detailed. 00:08:10.140 --> 00:08:13.160 But it talks about how I worked on Python Bytes and Talk Python Web Platform. 00:08:13.420 --> 00:08:14.260 And what did I do? 00:08:14.700 --> 00:08:15.060 I don't know. 00:08:15.120 --> 00:08:16.920 I worked on some admin backend stuff. 00:08:17.160 --> 00:08:20.760 There's a really neat admin section that no one is going to ever see. 00:08:21.120 --> 00:08:22.340 A few people will ever see. 00:08:22.540 --> 00:08:22.720 Yeah. 00:08:23.640 --> 00:08:26.360 Because, you know, it's private to make the whole thing go. 00:08:26.740 --> 00:08:29.620 Some DevOps seeing the Sentry thing coming back again. 00:08:29.740 --> 00:08:31.960 But it gives you also like really nice graphs. 00:08:32.140 --> 00:08:32.320 Too neat. 00:08:32.760 --> 00:08:33.320 What you wanted. 00:08:33.719 --> 00:08:35.900 Feature implementations, bug fixes, UI styling. 00:08:36.219 --> 00:08:39.380 And apparently I wanted features most and then bug fixing. 00:08:39.839 --> 00:08:40.460 That's good. 00:08:40.700 --> 00:08:45.600 The languages that I interacted with through Claude, the type of sessions, the tools. 00:08:46.140 --> 00:08:47.580 I use 2,680. 00:08:48.300 --> 00:08:49.520 That's a jump on mine too. 00:08:50.620 --> 00:08:50.860 Nice. 00:08:51.060 --> 00:08:51.160 Awesome. 00:08:51.600 --> 00:08:51.760 Yeah. 00:08:51.880 --> 00:08:57.260 And then, yeah, it goes and talks about how you operate like this techno and such and such and such and so on. 00:08:57.520 --> 00:08:58.220 I think this is great. 00:08:58.800 --> 00:09:03.740 I mean, if you're a heavy Claude Code user, does it also track the co-work sessions or is it only Claude Code? 00:09:04.020 --> 00:09:06.600 I think it's only Claude Code. 00:09:06.720 --> 00:09:09.400 It depends if it stores it as a session. 00:09:09.860 --> 00:09:10.440 I think it does. 00:09:10.560 --> 00:09:15.040 Well, at least I know the agency picks up co-work and Claude Code sessions. 00:09:15.340 --> 00:09:20.580 I would imagine it would because it just, all this thing does is it just reads and analyzes your session folder. 00:09:20.920 --> 00:09:21.040 Right. 00:09:22.100 --> 00:09:28.040 The types of errors that were encountered, stuff that works, but also stuff that went wrong. 00:09:28.240 --> 00:09:37.880 So here's where it gets to be like, you can actually make buggy intermediate code or self-inflicted tool collisions, claiming correctness when it shouldn't have wrong, wrong target. 00:09:38.300 --> 00:09:39.940 And then here's the last thing I'm going to talk about. 00:09:40.140 --> 00:09:40.920 And you guys can check. 00:09:41.380 --> 00:09:42.480 That's exactly what I was going to ask about. 00:09:43.360 --> 00:09:44.180 It looks like it's right here. 00:09:44.900 --> 00:09:45.060 Yeah. 00:09:45.140 --> 00:09:48.480 Here's at the bottom, it says existing Claude Code features to try. 00:09:48.980 --> 00:09:50.400 And you can just say copy all. 00:09:50.870 --> 00:09:54.980 And it literally says paste this into your cloud to empty. 00:09:55.290 --> 00:09:58.480 And you could just check and uncheck whichever you want and copy and paste. 00:09:58.800 --> 00:10:09.460 So it says just give you little things that you can record for Claude Code that will make it better based on its sort of, it's like a profiler guided optimization sort of thing. 00:10:10.020 --> 00:10:13.180 But for AI, it says verification before claiming done. 00:10:13.560 --> 00:10:14.900 This sounds like a superpower sort of thing. 00:10:15.600 --> 00:10:28.200 never claim a layout UI change is correct based on reading code take a headless screenshot with Playwright and browser tooling inspect it before you bother the person and tell them it's done because they often have to give you screenshots back and say actually it's not done 00:10:28.540 --> 00:10:42.339 editing rules, sentry triage, workflows and then finally, this is a cool part too, is custom skills why don't you copy this and it'll tell you you like this sentry triage thing? Copy this, paste it in there 00:10:42.300 --> 00:10:46.340 and Claude will create a skill, which then you can just say slash sentry triage. 00:10:46.700 --> 00:10:49.880 And hooks you can put in, all kinds of stuff, new ways to use it. 00:10:49.980 --> 00:10:53.460 Like there is a ton of stuff in here that is proactive, right? 00:10:53.660 --> 00:10:58.020 And then it gives you a little, I don't know if it's inspirational or whatever, but just like a call, a pull quote. 00:11:02.220 --> 00:11:03.480 So what do you think of this, Calvin? 00:11:04.220 --> 00:11:04.820 It's pretty cool. 00:11:05.040 --> 00:11:07.180 I think more folks should be using these kinds of tools. 00:11:07.360 --> 00:11:11.860 Like if you are a Claude code junkie, like this is perfect for improving your game. 00:11:12.480 --> 00:11:18.980 It's a feedback loop right into the human side of things that you get to now feed back into the Claude side of things to improve your experience. 00:11:20.280 --> 00:11:27.840 To me, folks are probably still just trying to one-shot a thing without enough guidance to the agents, ending up in kind of weird spots. 00:11:28.260 --> 00:11:30.440 It's all about honing and shaping your experience. 00:11:30.920 --> 00:11:36.040 Like those Claude MD additions, very specific to you, like the kinds of work you're doing. 00:11:36.060 --> 00:11:38.880 You don't want to overload that file with too much stuff. 00:11:38.980 --> 00:11:42.640 you want it to be very specific to the kind of workloads you're doing. So I love it. 00:11:43.859 --> 00:11:46.980 Guide people to the right place. That's right. It also 00:11:47.120 --> 00:12:01.040 talks a lot about your AI engineering style, if such a thing exists. You work in phases instead of just one shot. Plans, phases at a time, verification gates, adversarial sub-agents, and all these kinds 00:12:01.040 --> 00:12:06.959 of things that you might use or don't use, but probably should be. So I think it's... 00:12:06.980 --> 00:12:09.640 It has a section called Impressive Things You Did. 00:12:09.820 --> 00:12:11.400 So, you know, give yourself a little warning. 00:12:11.480 --> 00:12:14.180 Yeah, a little pat on the back, feel pretty good about yourself there. 00:12:15.600 --> 00:12:16.460 Absolutely, absolutely. 00:12:17.120 --> 00:12:17.620 Very cool. 00:12:17.920 --> 00:12:19.040 Well, you know what else we can feel good about? 00:12:19.400 --> 00:12:19.840 What's that? 00:12:20.060 --> 00:12:24.060 We now have post-quantum cryptography in Python. 00:12:24.500 --> 00:12:30.980 So Python cryptography in version 48, which actually was back in May, but I don't believe we ever reported on this. 00:12:31.360 --> 00:12:37.040 but it is shipping two variants of some post-quantum standard enablements. 00:12:37.220 --> 00:12:41.060 So we've got MLKEM, which is key encapsulation. 00:12:41.460 --> 00:12:45.120 So basically the way we do it now with Diffie-Hellman, it's a key exchange. 00:12:45.520 --> 00:12:47.300 We share bits and pieces and parts. 00:12:47.920 --> 00:12:50.740 Key encapsulation actually shares the whole key at once. 00:12:51.120 --> 00:12:52.740 So it's a different form of doing this. 00:12:52.990 --> 00:12:55.580 And then MLDSA, which is the signatures. 00:12:56.400 --> 00:12:59.260 These are all, again, part of the NIST post-quantum standards. 00:13:00.060 --> 00:13:01.100 Why is this a big deal? 00:13:01.260 --> 00:13:01.900 Why do we care? 00:13:02.490 --> 00:13:17.140 This all tracks back to a executive order from our White House, the June 22nd executive order setting the federal deadlines for post-quantum key establishment by the end of 2030 and post-quantum signatures by the end of 2031. 00:13:18.120 --> 00:13:23.320 Trail of Bits here did a great write-up on what it's all about, what you're getting for it. 00:13:23.720 --> 00:13:32.320 There's actually an interesting article if you go read here from 2024 about quantum resistance isn't the main benefit of post-quantum code. 00:13:32.460 --> 00:13:37.160 There's actually some benefit to the new algorithms that were developed as part of this. 00:13:37.540 --> 00:13:41.740 It doesn't come without some difficulties or drop-in. 00:13:41.780 --> 00:13:43.160 It's not a drop-in swap. 00:13:43.480 --> 00:13:56.160 If you're using the standard ED25519 or the standard Diffie-Hellman key exchange pieces, some of them work alike, some of them don't. 00:13:56.460 --> 00:14:09.740 For example, the key signatures or the encapsulate versus decapsulate piece rather than the key exchange isn't the same, but they're going to be putting in the SLHDSA, all these acronyms that your mind can kind of like 00:14:10.460 --> 00:14:24.799 explode on. But the wire sizes on the wire size of the signatures and key encapsulations are much, much bigger. So for example, if you're using the ED25519, that's 64 bytes on the wire for the 00:14:24.940 --> 00:14:34.980 the key exchange versus 3,309 for the MLDSA signatures. So it's probably negligible even 00:14:35.300 --> 00:14:46.120 today on our current hardware. Hardware from a few years back would have been having noticeable time processing these signatures. New hardware should see even less of this being an impact on performance. 00:14:46.380 --> 00:14:57.740 So there's more of a processing performance impact to swapping over to these new post-quantum signatures and the key exchanges. But it's also a big deal because, let me go over here into 00:14:58.240 --> 00:15:11.100 showing you the PyPI downloads. Cryptography is number 10 in the past day, number 10 in the past week, number 10 in the past month. So since this is downloaded very, very often, we're on version 00:15:11.500 --> 00:15:26.100 50 now, people have the capability to upgrade their use of this encryption to that post-quantum capable world. Why do you care? Basically, there are malicious actors out there that are 00:15:26.980 --> 00:15:31.940 hoarding currently encrypted data, waiting for a time when these quantum computers become a reality. 00:15:32.640 --> 00:15:46.859 It's still an if. We don't know that to be for certain that they'll ever reach a capability to be able to decrypt these things. But quantum would unlock the ability to decrypt all the traffic that has been basically passed across the wire. Most traffic has been passed across the wires since 00:15:46.880 --> 00:15:56.520 the internet and networks were a thing anytime it was encrypted. I can tell you a lot of places are already using the post-quantum encryption internally. For example, I know Amazon has 00:15:56.520 --> 00:16:08.980 been a big backer of some of the post-quantum enhancements to requests in the HTTP libraries so that all their internal operations behind the scenes in AWS Cloud are using post-quantum 00:16:09.320 --> 00:16:24.060 encryption. This means that if someone does hoard or stockpile your data that has been used with these post-quantum tools that they shouldn't or they'll have a much greater difficulty actually decrypting that data so it's important here to know that you should be on board with uh you know 00:16:24.160 --> 00:16:38.260 maybe not this week but maybe in the future very soon using these new versions of these key exchanges and signatures so that you'll be ready for the post-quantum world i guess the protocols haven't caught up yet but usually you know you're not going to be running a post-quantum cert bot 00:16:38.460 --> 00:16:49.580 this week but it'll be coming very very soon and everyone basically has it in their hands that's pretty wild yeah it feels like the future but i know if you look i mean the pi pi pi boy people 00:16:49.580 --> 00:16:55.980 are going to get on me about saying pi pi um it's pi pi and the downloads for i think it's worth a 00:16:56.220 --> 00:17:10.079 sidebar here uh is it when i first started doing podcasts talk python and python by yeah especially talk python people will come on and they would say pi pi frequently and obviously there's the 00:17:10.480 --> 00:17:12.620 And that's a fair pronunciation of it. 00:17:12.689 --> 00:17:17.480 But there's also PYPY, which is the runtime, which is the JIT version. 00:17:18.040 --> 00:17:21.180 And there's all this – it was pretty evenly split. 00:17:21.270 --> 00:17:23.339 And I'm like, you know, this is – I need to know. 00:17:24.069 --> 00:17:31.700 You know, inquiring minds, as they say, from the checkout line investigators, I guess. 00:17:32.070 --> 00:17:32.980 Yeah, exactly. 00:17:34.700 --> 00:17:40.040 So I asked a lot of the folks that work at PyPI, how do you pronounce this? 00:17:40.340 --> 00:17:41.600 Like, you guys made it. 00:17:41.900 --> 00:17:42.900 You can kind of pick the name. 00:17:42.980 --> 00:17:44.200 Like, what would you say, right? 00:17:45.120 --> 00:17:48.360 And Donald stuffed and some of the other folks are like, PyPI. 00:17:48.780 --> 00:17:50.620 And I asked Guido and he said, PyPI. 00:17:50.680 --> 00:17:51.120 I'm like, all right. 00:17:52.240 --> 00:17:53.540 They have it in their FAQ. 00:17:53.800 --> 00:17:55.880 It's been pointed to me by some of our listeners. 00:17:56.700 --> 00:17:59.380 That is the official designation. 00:18:01.060 --> 00:18:02.220 Maybe it's just generational. 00:18:02.280 --> 00:18:06.420 I mean, when you and I were growing up, it was called the cheese shop at the time. 00:18:06.620 --> 00:18:08.840 So sorry for the PyPI diversion. 00:18:09.780 --> 00:18:13.140 I'm working hard to break some pronunciation habits. 00:18:13.720 --> 00:18:14.300 And that's one of them. 00:18:14.580 --> 00:18:15.460 No, it's amazing. 00:18:15.640 --> 00:18:15.840 Keep going. 00:18:16.480 --> 00:18:18.720 I thought a little bit of history might be fun for people. 00:18:18.900 --> 00:18:19.680 It is a little fun. 00:18:19.760 --> 00:18:20.940 It is fun, but it is PyPI. 00:18:21.620 --> 00:18:22.840 But cryptography, you can see here. 00:18:22.900 --> 00:18:24.140 I mean, this is incredible. 00:18:24.460 --> 00:18:33.420 The fact that this work got sponsored, got implemented, and now is immediately in the hands that everyone needed to be able to do the right thing. 00:18:33.820 --> 00:18:35.600 This sets us up in a really, really good spot. 00:18:37.120 --> 00:18:41.700 Yeah, again, this was funded by the Sovereign Tech Agency, and I had to go look up who that was. 00:18:41.920 --> 00:18:48.240 So if you check that out, basically, it's a German industry or agency that put forth the money to make sure this all happens. 00:18:48.700 --> 00:18:51.580 And as they say here, the engine room for digital sovereignty. 00:18:51.980 --> 00:19:02.120 And so I thought it was cool to give them a shout out for sponsoring it and then for Trail of Bits doing the work, putting in place the Rust bindings, the cross backend APIs, like all the tests, everything is in there. 00:19:02.480 --> 00:19:08.040 So I think it's a huge win for the Python community that we are ready to go for this post-quantum world. 00:19:08.300 --> 00:19:09.660 I do too. I think it's totally neat. 00:19:09.760 --> 00:19:24.580 Also, I think it's neat that it was put into cryptography, as you pointed out, the 10th most popular downloaded API thing, rather than here's a Python-quantum-crypto whatever, right, that nobody uses. 00:19:25.040 --> 00:19:34.000 Because I think a lot of people don't realize how significant the default transitive dependency thing is, right? 00:19:34.180 --> 00:19:39.580 Like we would all like to switch from this, but seven of the 10 dependencies I have all depend on that. 00:19:40.050 --> 00:19:43.880 And I've just got to take that because that's what they use, not what, you know what I mean? 00:19:43.890 --> 00:19:46.680 And so I think this will just like come along to a lot of projects. 00:19:47.020 --> 00:19:48.060 Yeah, this is a real good model. 00:19:48.280 --> 00:19:55.420 I also love the fact that it shows that community bridging with the Rust community on it as well. 00:19:55.860 --> 00:19:58.040 And so Python has really great support for bringing in. 00:19:58.280 --> 00:20:01.860 It used to be we were using C extensions for things that required performance like this. 00:20:01.980 --> 00:20:04.540 And now we've really got a native way to use Rust. 00:20:04.840 --> 00:20:05.860 It's been in a while. 00:20:06.260 --> 00:20:07.440 This is a great example of that. 00:20:07.640 --> 00:20:08.600 Yeah, it definitely is. 00:20:08.700 --> 00:20:09.220 That's super cool. 00:20:09.480 --> 00:20:09.700 All right. 00:20:10.020 --> 00:20:11.460 Well, Michael, let's move on. 00:20:11.500 --> 00:20:12.920 We got a sponsor segment. 00:20:13.300 --> 00:20:15.200 We do have a sponsor segment. 00:20:15.480 --> 00:20:19.660 So let me quickly tell everyone about XWeather from Vysala. 00:20:20.040 --> 00:20:22.680 And so they're supporting the show last week and this week. 00:20:22.900 --> 00:20:24.500 And they've got a really cool project. 00:20:24.840 --> 00:20:36.720 So if you're using agents that can write code, summarize documents, automate workflows, but they don't know about the weather, awareness of the world around them, well, you should check out XWeather, all one word. 00:20:37.060 --> 00:20:39.380 This is where today's sponsored XWeather comes in. 00:20:39.700 --> 00:20:50.120 It combines enterprise-grade weather intelligence with agent-ready APIs, natural language capabilities, and an MCP server built for tools like Cloud, Codex, Copilot, and Modern IDs. 00:20:50.720 --> 00:20:57.060 So your agents can adopt workflows, automate responses, and make better decisions based on real-world conditions. 00:20:57.560 --> 00:21:01.840 Backed by Vaisala, whose instruments fly on NASA missions to Mars. 00:21:02.340 --> 00:21:02.960 That's super cool. 00:21:03.200 --> 00:21:04.480 Talk about the weather there. 00:21:05.080 --> 00:21:08.060 Our weather API works on multiple planets. 00:21:08.240 --> 00:21:08.620 That's pretty awesome. 00:21:09.460 --> 00:21:17.940 So they deliver trusted data and unique insights that go beyond conditions to actual impact, like real-time lightning strikes, road service forecasts, and so on. 00:21:18.300 --> 00:21:23.420 So start with 15,000 free API calls a month and pay only for what you use as you grow. 00:21:23.780 --> 00:21:26.820 Xweather is a full weather stack for developers by developers. 00:21:26.970 --> 00:21:31.280 So start building for free today at pythonbytes.fm/xweather. 00:21:31.640 --> 00:21:34.220 The link is your podcast player show notes and on the episode page. 00:21:34.460 --> 00:21:37.020 Thank you so much to xweather for supporting Python Bytes. 00:21:37.130 --> 00:21:37.980 We really appreciate it. 00:21:38.120 --> 00:21:41.280 Yeah, it's amazing how much weather, how weather can impact all of our lives. 00:21:41.540 --> 00:21:46.160 So having a tool like this to be able to check in and get high quality weather results. 00:21:46.680 --> 00:21:47.240 I love it. 00:21:47.430 --> 00:21:47.760 Super awesome. 00:21:48.220 --> 00:21:49.280 Yeah, yeah. Very cool. Very cool. 00:21:49.500 --> 00:21:50.720 Well, thanks to them for sponsoring. 00:21:51.180 --> 00:21:54.100 And up next, we got some MCP news. 00:21:54.520 --> 00:21:57.440 So the model context protocol has gone stateless. 00:21:58.380 --> 00:22:04.380 The folks over here at RealPython, thanks to Philip for giving a good roundup on the news around this. 00:22:04.720 --> 00:22:14.980 But not only has the MCP spec gone stateless, so if you check out the 2026.07.28 spec, it landed last month. 00:22:15.400 --> 00:22:18.160 And the Python SDK also shipped a v2. 00:22:18.740 --> 00:22:21.240 That is the biggest rewrite since MCP launched. 00:22:21.580 --> 00:22:27.800 And there's a big breaking or not a breaking change, but a forward moving change here, which is that going stateless. 00:22:28.180 --> 00:22:37.360 Prior to this new version of the MCP spec, you would establish a session with your MCP servers from your clients and hold that session open kind of like WebSockets. 00:22:37.620 --> 00:22:40.560 But that can be it has obviously troubles with scaling. 00:22:40.960 --> 00:22:49.100 If you are trying to hold a big session open and have many, many clients, you're going to need a lot of resources to be able to handle a ton of clients against your MCP server. 00:22:49.580 --> 00:22:56.360 The stateless version of this is we now can basically send over enough context information to headers. 00:22:56.640 --> 00:23:01.300 There's a new initialize and initialized handshake and the MCP session ID headers. 00:23:02.560 --> 00:23:05.080 The MCP session ID header has been retired. 00:23:05.600 --> 00:23:19.800 And now you basically make those requests and you can optionally do what's called the round trip, multi-round trip request to the MCP server if you need to pass in more context or additional information. 00:23:20.260 --> 00:23:23.360 So this was a big deal for the MCP world. 00:23:23.780 --> 00:23:29.180 Luckily, the MCP servers are backward compatible to the 1.0 version of this. 00:23:29.620 --> 00:23:34.800 So for us as Python users, this is actually more important because we get the Python SDK 2.0. 00:23:35.360 --> 00:23:39.140 Biggest change here, some folks may recognize it as Fast MCP. 00:23:39.450 --> 00:23:41.300 It is now just called MCP Server. 00:23:42.000 --> 00:23:44.960 There's no deprecation or backward compatibility piece of that. 00:23:45.010 --> 00:23:51.980 So if you are writing an MCP server based on Fast MCP, you're going to need to go basically fix that code before you release your new versions. 00:23:52.030 --> 00:23:57.920 If you want to use the Python SDK version 2.0, lots of cool changes or fixes in there. 00:23:58.420 --> 00:24:06.620 Another big piece of the new spec, if you haven't used it, which is there's tasks that were experimental and apps that were experimental. 00:24:06.940 --> 00:24:17.180 And if you've not used apps, it's kind of neat because it allows apps to send back an HTML interactive sandboxed mini application as part of the MCP specification. 00:24:17.880 --> 00:24:18.680 Again, it was experimental. 00:24:18.720 --> 00:24:20.020 We didn't know where that would land. 00:24:20.360 --> 00:24:22.160 That has been picked up and officially put into the spec. 00:24:22.520 --> 00:24:26.560 The SDK 2.0 doesn't have that included just yet, but it's coming. 00:24:26.880 --> 00:24:39.340 So if you want to be able to use those sandbox interactive HTML pieces, that is going to be really nice because you basically, if you ever had an MCP server for the weather, for example, that it could actually return UI to the end user of that MCP service. 00:24:39.470 --> 00:24:49.060 If you were using it in Cowork or Claude through the web API or web version of this, you can now get rich interactives as part of the new spec. 00:24:50.060 --> 00:24:52.400 Pretty cool because basically, yeah, we're good to go. 00:24:52.800 --> 00:24:56.600 If you're only calling MCP servers, this really doesn't matter mostly to you. 00:24:56.600 --> 00:24:57.580 You get the benefits for free. 00:24:57.980 --> 00:25:01.880 If you're shipping MCP servers, you know what your week's going to look like. 00:25:01.890 --> 00:25:08.760 You're going to be doing some subtle code changes and upgrades to get the latest versions of the SDK in place for you. 00:25:09.260 --> 00:25:09.480 I don't know. 00:25:09.630 --> 00:25:10.040 Have you written? 00:25:10.520 --> 00:25:12.700 Actually, Michael, you've written quite a few MCP servers. 00:25:12.980 --> 00:25:13.400 I have. 00:25:13.920 --> 00:25:16.860 You're rocking my world here, Calvin. 00:25:18.700 --> 00:25:22.140 So it sounds to me, let me lay this back and you tell me if this is correct. 00:25:22.340 --> 00:25:23.780 I'll do this on behalf of the listeners as well. 00:25:23.900 --> 00:25:28.900 So for those who don't know, MCP is not just a library or something like that. 00:25:29.060 --> 00:25:37.360 It's a standard or a protocol come up with by Anthropic and adopted by, as you've heard from our sponsor, and you're going to hear again in just a minute. 00:25:37.360 --> 00:25:37.540 Pretty much everybody. 00:25:38.200 --> 00:25:43.160 Pretty much everybody who cares about being friendly with having agents like Sentry, for example. 00:25:43.600 --> 00:25:49.360 So they've changed the specification and then Fast MCP has said, we're on it. 00:25:49.720 --> 00:25:51.880 And they adapted that to actually adopt. 00:25:52.100 --> 00:26:03.940 Yes. Okay. Yep. Yep. So if you go over here to the latest release, you will see that two weeks ago, we had the version two release to get you off of fast MCP into just MCP and MCP server. There's 00:26:04.080 --> 00:26:16.600 more types. The one client object replaces the old transport plus client session plus initialized stack. They also upgraded from HTTPX to HTTPX2. So you get a lot of benefits there and performance. 00:26:17.210 --> 00:26:25.560 I mean, this is all about performance and scaling. If we want to be able to scale the usage of these kinds of services. That's why HTTP has scaled so well for all of us because it's been stateless. 00:26:25.960 --> 00:26:30.060 It didn't require a lot of server resources. This gets us back into that scaling world for 00:26:30.320 --> 00:26:41.260 MCP style calls. Yeah, it opens up possible deploying models that are really tough. Like can you do web sockets to Lambda? It's not so easy, maybe with a proxy or something. 00:26:41.590 --> 00:26:53.480 Yeah, actually there's a little diagram in here kind of shows you the before and after. Before as the session was established. 00:26:54.000 --> 00:26:58.700 Here now they can just go to any number of session or any number of handlers that would be on the backend. 00:26:59.160 --> 00:27:07.560 So now you can set up a nice cluster with a much dumber load balancer, a less smart load balancer in front that can just route traffic. 00:27:08.020 --> 00:27:14.060 And again, those session pieces all stack on each other from a performance standpoint. 00:27:14.740 --> 00:27:18.120 So freeing up to be stateless is a huge impact on performance. 00:27:18.640 --> 00:27:19.880 Yeah, very neat, very neat. 00:27:20.020 --> 00:27:20.620 I'm excited for it. 00:27:20.840 --> 00:27:25.500 Yeah, yeah. I'm looking forward to leveraging this and writing some more clients that go stateless. 00:27:25.600 --> 00:27:31.200 So if you're using this in your production agentic workflows, this actually keeps MCP very viable. 00:27:31.660 --> 00:27:34.200 Yeah, very cool. It definitely does make it easier to work with. 00:27:34.580 --> 00:27:40.320 Yeah, you can even with the 1.0, you can disable streaming, but it's not quite as good. 00:27:40.580 --> 00:27:42.300 So this kind of formalizes that. 00:27:42.600 --> 00:27:45.800 Yeah, yeah. So that's the exciting news in the MCP world. 00:27:46.200 --> 00:27:50.540 I want to thank again the RealPython folks for giving us that bit of news. 00:27:52.000 --> 00:27:56.820 actually let's head on to yours Michael you got some more developer tool here it looks like 00:27:56.820 --> 00:28:11.220 I don't know I feel like somehow I got a hold of this and this should be all it should be all you so again a little history a lot of people call what happens when you type something and then you hit dot and like a list of options come up call that autocomplete 00:28:12.019 --> 00:28:19.400 in Microsoft developer land this is called IntelliSense as in intelligence sensing I don't know it's been around forever 00:28:19.580 --> 00:28:25.040 I was never a Visual Studio user, so I never used IntelliSense proper. 00:28:25.220 --> 00:28:31.260