Does AI Change the Government for Good? (Consequential)
November 8, 2024
Parker dives into the latest AI tooling and real-world workflows this week, highlighting Zed IDE, a powerful new prompt ecosystem, Warp terminals, and hands-on agent work—plus what the current geopolitical landscape means for AI tooling.
Zed IDE overview and impressions#
- A fast, Rust-based IDE aiming to replace traditional editors with AI-assisted capabilities.
- Core strengths:
- Performance: built in Rust, designed to be lag-free with multi-buffer editing (up to 120 FPS).
- Open source with a readable repo and a familiar feel for Vim/Emacs users.
- Strong project-wide capabilities: outline view, fuzzy search, and a dedicated AI panel.
- Notable features:
- Inline AI assistant (control/enter) for on-the-fly changes.
- AI contexts: shareable conversations that can be loaded into new sessions.
- Prompt Library: a built-in catalog of prompts you can reuse across projects.
- Zai model: new in-editor model tuned from Claude 3.5 Sonet, currently free.
- New workflow ideas: contexts, conversations, and in-editor prompts are integrated directly into the IDE.
- Notable UX notes:
- The AI assistant outputs aren’t applied automatically; you typically copy/apply results via keyboard commands.
- Collaboration panel feels experimental and may not suit everyone; it’s built into the editor as a chat-like space for bug discussion.
- Diagnostics pane shows multi-file errors in a dedicated view, helping you stay in flow without chasing errors across windows.
- Practical takeaways:
- If you code frequently with AI help, Zed’s in-editor contexts, prompt library, and multi-buffer UX can significantly speed up workflows.
- Expect small gaps between assistant outputs and applying changes; there’s a learning curve to the precise apply-flow.
In-editor AI workflow: contexts, prompts, and commands#
- Contexts vs. conversations:
- Contexts store past conversations and project references; you can load them into new sessions.
- Conversations provide the memory you reference in subsequent AI interactions.
- Prompt Library:
- A centralized library of prompts you can reuse anywhere in the editor.
- Improves consistency and speed by avoiding re-creating prompts for similar tasks.
- Commands and navigation:
- Command Palette: Forge slash commands for quick actions (insert file, fetch docs, etc.).
- File juggling: fetch docs or external sources and insert them into your current file as code fences for easy editing.
- Project navigation and outline: easy switches between files and project structure without losing context.
- Quick tips:
- Use the inline assistant for on-page edits; use the right-side AI panel for broader context and historical prompts.
- You can load a prompt or context into a session to steer the AI during large changes.
The big prompt library and model integration#
- What it is:
- A curated collection of system prompts and LM instructions from popular models (e.g., Claude, OpenAI, etc.), bundled for quick reference.
- Why it matters:
- Prompt engineering remains one of the biggest levers for AI quality; having examples handy helps you blueprint better prompts quickly.
- Parker’s take:
- The library is a solid reference, but you’ll still need to tailor prompts to your exact use case. It’s a great learning and baseline tool.
Bolt vs. in-editor control vs. full-stack automation#
- Bolt concept:
- A browser-based tool that can generate full-stack results (e.g., a Spotify clone) and manage dependencies.
- Trade-offs:
- Great for beginners or quick scaffolds.
- Can produce “spaghetti code” or less-tweakable outputs; you lose some control over fine-grained architecture.
- In-editor workflows typically offer more control and better integration with your project’s context.
- Takeaway:
- Use Bolt for rapid scaffolding or learning prompts, but rely on in-editor tooling for fine-grained control and maintainability.
Warp terminal: natural language, faster commands#
- What it does:
- Natural language-enabled terminal, with a context-aware toggle that activates when you start typing non-shell commands.
- Why it’s useful:
- Speeds up common tasks (migrations, searches across docs, quick scripting) by letting you describe what you want in plain language.
- Context handling means you don’t have to repeat everything in every session.
- Practical tip:
- Leverage Warp to prototype quick commands and then drop into a shell for exact scripting once you confirm behavior.
Geopolitics and AI: Anthropic, Palantir, AWS, and the government angle#
- Key development:
- Anthropic signs a significant collaboration with Palantir and AWS to provide AI-enabled insights and capabilities to U.S. defense and intelligence workflows.
- Implications for developers:
- Public-sector and defense-adjacent AI tooling influence which models and providers gain traction.
- Expect shifts in procurement, data governance, and model access that could steer tool choices in enterprise settings.
- Perspective:
- The landscape is moving toward a mix of private-sector AI power and government-backed deployment, with notable attention on regulatory capture risk and ethical guardrails.
- Quick take:
- Keep an eye on XAI (the new AI initiatives from around the same time) and Claude/01-era progress; government access to compute and data can reshape timelines and tooling affinities.
Claude, haou, 01, and the compute race#
- Claude updates:
- Claude haou (new model) announced; Parker didn’t find it materially helpful in his use case, but acknowledges ongoing model iterations.
- 01 and the compute race:
- Anticipation around a new 01 model drop; early previews are watched as a leading indicator for capabilities.
- XAI’s rapid build-out (reportedly faster than typical timelines) is noteworthy; watch for industrial-scale compute availability by December.
- Takeaway:
- Model updates are frequent; stay on the lookout for capabilities that directly impact your workflow (e.g., chain-of-thought reasoning, code generation, safety features).
Agents and a real-world site audit workflow#
- Core idea:
- Agents are specialized, task-focused components that perform a specific job well; multi-agent workflows beat trying to cram everything into a single agent.
- The site audit agent (demo project):
- What it does:
- Visual testing: captures screenshots across desktop/tablet/phone sizes for all views.
- Performance: runs Lighthouse/PageSpeed checks.
- SEO and accessibility: analyzes domain/page authority, accessibility semantics, and SEO best practices.
- Security checks: basic policy and SSL checks.
- Framework detection: identifies underlying tech stack (to reproduce or learn from the stack).
- How it works (high level):
- The agent orchestrates multiple tools via a CLI. You pass a URL (or multiple URLs) and pick which checks to run.
- Outputs are aggregated into a large JSON, then summarized and turned into prompts for an LM (Claude in this case) to generate both a client-facing report and a developer-oriented improvement plan.
- UI and scale:
- Headless by default; outputs can be organized per-URL with folders and test results.
- The workflow emphasizes chunking long lists so the LM can handle the data without hitting token limits.
- Why this matters:
- This is a concrete example of multi-tool orchestration with AI, showing how to produce actionable, comprehensive audit reports in minutes rather than hours.
- Takeaway:
- For consultants and product teams, building modular agents for distinct tasks (visual QA, performance, SEO, security) can dramatically speed up audits and enable scalable reporting.
Practical takeaways and actionables#
- Try Zed if you want a high-performance, AI-enabled editor with native prompts and contexts; watch for the friction between outputs and applying changes.
- Leverage the Prompt Library and Contexts to accelerate repetitive AI tasks across projects.
- Use Warp or similar natural-language terminals to speed up shell work and reduce boilerplate typing.
- Explore multi-agent workflows for complex tasks (e.g., site audits, bug triage, accessibility checks) to reduce manual labor and improve consistency.
- Stay informed about government-facing AI partnerships and large-model compute launches; they shape which providers and models become mainstream in enterprise settings.
Next steps and what to watch#
- Keep an eye on Claude 01 and XAI’s compute roadmap for real-world impact on coding and AI-assisted workflows.
- Watch for deeper integration of the AI assistant in editors (changes that can be applied directly from prompts) and improved UI flows.
- Monitor how AI governance and procurement shifts influence tooling choices in teams and startups alike.
Links#
- Zed IDE - Rust-based, open source editor
- Zed AI - In-editor AI model used by Zed
- Claude by Anthropic - Claude 3.5 Sonnet and newer variants
- Anthropic + Palantir + AWS collaboration - Government-focused AI deployment
- Warp terminal - Natural language terminal
- Wappalyzer - Chrome extension for framework detection
- Google Lighthouse - Performance testing tool
If you want deeper dives on any of these topics, tell me which segment you’d like expanded into a full tutorial or hands-on walkthrough.
Transcript
[Music] all right it's a new week it is November 8th and that means that there's a lot of new news out a lot of stuff happened in AI this week we're going to cover all of it sit tight let's get right into it so the first one I want to talk about is actually a potential for a new IDE a potential for a tool that could just totally disrupt the way that people are used to coding now if you haven't heard of it it's called Zed and Zed is made by the founders of Adam so Adam was a really popular open source IDE from way back in the day you may or may not be familiar with it but there's been a new one that they've been working on called Zed and it's really interesting not only from the AI perspective but just the performance of it so one of the biggest Pros is it's written in Rust which means that it's going to be very fast there's no lag you can look at the main page on it where you know it's like written from scratch and rust it's totally open source you can actually go to the repo and read through it and learn a whole lot from it has all the llms in it and it's using a different pattern and this week was the first week of me trying to use it full-time in place of cursor cursor is great it's really popular but it's very you know the thing just crashes a lot it has a lot of trouble they Fork vs code they're putting a lot of stuff into that the composer it it's great for really small pieces of work but I find that if you're trying to feed it too much context that it actually starts to really struggle another piece about it is it's collaborative we'll talk about that but on the AI side actually it kind of flips the script on how you're used to doing things so instead of you having this composer and jamming a bunch of stuff in there and then having it rip you actually do things a little bit differently so we'll touch on that in a second but it's just so fast it's I really can't get over that part and it has the ability to have these multi buffers so it's running at 120 frames per second and then if you wanted to change something you can open multiple buffers in the same file they use the word buffer a lot which is great it's very close to them or emac and you can make changes across all these places really fast there's also the outline panel so if I'm in a project and this is a a note taking app and if I'm in a project and I want to open up what this uh file has in it I can do command shift B and then it'll open up the outline if I want to do the normal search it's command shift F and I can search the project with fuzzy search and then on the right hand side you have this AI pan so you'll notice it's really similar actually to using a playground which I think is a a great idea so if I open up a file let's just find one um maybe this index file and it has the same hot Keys VSS code so they're using a lot of the ones so you don't have to totally rethink your whole world when it comes to IDs I can just take a look and if I want to do something in line and I can grab the text and I can hit uh you know command or sorry control and enter and I can do a transform there and I can actually see what are the tokens on this and then I can have it do the work so there's the endline stuff which we're used to but then what's different is here on the right you have these contexts is what they call them and there's a key difference here it was a little confusing for me to to understand when I first was going through the assistant stuff but you know you have this AI assistant panel so you can create the chats manage interactions y y y you have the inline assistant which we just talked about with just control enter Then you can actually do all the configuration you want to do and then context so contexts are really just like past conversations that you've had that can then be loaded into new conversations so they make a distinct uh difference there it's like contexts are like conversations it's a collaborative tool for sharing information between you your project and the model and and the model can reference content from your active context in the assistant panel but elsewhere like the inline assistant so these are all saved and again all fast I've noticed with cursor I have to go back and I'm deleting them to make sure that everything's still quick none of that's the case here so all those end up getting put in this history panel so on the right hand side when I do uh command r that opens up the window and the basic UI is I can either create a new context which is basically a conversation that has all the history I can look at the history so if I want to load one of these into the next session I can do that I have a prompt Library this is great this is one of the best features because I can just bring these in anywhere I can reference these all the time so instead of me having an obsidian doc with all of them or an obsidian folder or a bunch of text expanders I have my library right in here and I think this is one of the game changers because I can load all these in here again it's showing me all the stuff that I would expect to see as a developer in one of the playgrounds and like the token count I think that's really helpful so I can start putting all these in here they're searchable you can create more of them and then you have the configure session or uh view which this is a big one so they've introduced their own model and it's just called Zai and they're giving it out for free right now so this is another reason why I find it really helpful and they've taken the latest Claude 3.5 Sonet and tuned it it's crazy it's crazy good and it's free right now so that's a big deal and if you wanted to you can use something else but I'm not I'm using Zed which is just son at 35 tuned and that's really great so if I wanted to create a new context you'll see it's loading this is one thing that I was like kind of confused about where I tried to load one in but it doesn't doesn't seem to want to work but you have these commands and so if I do Forge slash then I'm immediately shown all the different commands that they have in here and they're adding more of them so you have insert file so if I wanted to work on the index. TSX which is on the left hand side there's a couple ways that I can get it in there I can either uh grab the file and then do command shift right arrow and put it in there and then what it's actually doing is it's putting that in a code fence so it's taking that it's put it in there and that's really helpful because then I can change it if I wanted to it's not just this tag and again it was really fast if I wanted to let's say um I wanted to open up a couple more windows like the login and the uh the pricing and privacy whatever then I can just do forward slash and then I can do file and then I can tag more of them if I wanted to do tab I can do Tab and that just takes the original one I can do Tab and then I can do tab all and then it brings in all of them again so it's really pretty helpful in that way where you're just doing it quick and you get fast so if I wanted to bring in documentation I can do a fetch so if I do fetch space and then let's say I had some stripe docks or something like uh let's just pull up stripe dock Bo one second get started with payments okay so I grabbed a link and then I can just hit that now it's bringing that in so it has all these really cool commands I can do diagnostics which is all the errors I can include warnings if I want I can do the specific path of the Diagnostics I want the terminal all all these are really helpful you can do now so it can know if you're doing date related things you can insert the prompt from the library that you had so you can start to see when you have these commands how powerful it really becomes so then if I wanted to let's say for example I want to say what could we change here on this index um by by the way that was the command pallet so that's command p and it's just opening it up very fast again if I wanted to do the settings I can do shift command p and then you can actually control the assistant you have all these different options in here it's really really really cool so if I just say on the index is that the right one um yeah if I just let's just do something simple like change the colors to dark mode now as I'm editing here when I hit enter I can just create more lines if I wanted to do more commands I can do those each command has to be on its own line if I wanted to insert that selection that was one of the things that I did earlier with the command shift right arrow then I just send that off and now it's cooking so I'm getting back the information from the assistant one thing that I would say is a con is most people are used to using the assistant on the right side and then having the ability just to kind of have it apply the change but this is not the case here so instead you're actually looking at it and then you can bring that in or what they recommend on the site is that then you would do command enter here and say apply the changes and bring in the context but you know I haven't gotten that to work like it says that if I do command question mark for context that it's going to bring that in but I don't know that's one of the one of the cons is I've been using it where I have it give me the output and then I'm just pasting it in there um I know that they had a feature in a previous version called workflow that would go and do it similar to how you would expect in composer but that's one thing that I I don't quite understand um just yet so it generates the changes there one thing that I find really helpful as well is if I start doing a conversation or a context and I start to get a response I can actually Escape it so if halfway through here I don't like what it's doing I can just hit escape and it's just done like it's cut off and that's not the case with cursor it's you know it'll start chugging along you might get the the beach ball and that allows you to steer this better so this whole thing is editable both the outputs and the inputs and I get the live update of the tokens it also renamed the context so that context always has a fresh name based on what we're talking about so if I wanted to bring this into the next chat then I could do so so I think that's helpful as well because I can kind of change history and how it's written so that's really cool I think that that's going to be helpful um one thing that is also good on the AI side is the diagnosis I kind of mentioned before but you actually get a screen that's dedicated to it so this is similar if you've used like X code um I think they do this in other IDs like in TJ but you have one screen for all the problems so it's not just this thing at the bottom that's a little panel it actually brings all them men and that's helpful because you can see from project level like what's going on where are the errors and then you can do that multifile buffer you can also do the outline view that will be dynamic to the errors that you do have and it just makes you move a little quicker and one of the things that might be controversial is this collab panel so a lot of people like I accidentally open this I'm like where am I and it's funny cuz this is basically like a zoom Channel like this is uh kind of like Discord they didn't want to do Discord they just built it in so if I wanted to I could join this and start talking to this guy Anders who's working on bugs I can also see the chat so there's people chatting about stuff and it's really funny because everyone's just confused about what this thing is and then there's notes that are assciated with the channel so because it's open source a lot of people are in here and you can learn a lot about how the product's being built so I don't know if this will stick but I think it's interesting and Es lent is one of the things that this gets confused by the Diagnostics so that's something to keep in mind but I think overall that this is going to be really interesting to watch uh I think I'm making the full switch I think the only thing that's blocking me on that is just the kind of the gap between the assistant panel and then making the changes but I found that the outputs were actually really really good and again it doesn't crash so that's there and then let's see what are we talking about next oh the big prompt Library so this is the second piece of news I want to cover this is a collection of promp system prompts and LM instructions and in here you just have a bunch of the different popular prompts so you can see when you come in here um this guy's basic basically like come up with a bunch of the different prompts you somehow figured them out some of them are older than others but you can see what the perplexity desktop system prompt is and I don't think you'd necessarily use this in place of these products but you can see how they do it like this is the cursor one so you can see how they're making it work uh same thing with the Microsoft co-pilot that one's a lot older they have anthropic effect they have open AIS perplexities this is just really cool to see cuz prompting is one of the biggest keys to uh you know making these things effective one of the things that a lot of people are finding is this bolt product is able to make clones of things really quickly now I think it's interesting but if you actually read the code the quality is you know it's controversial from what I've seen it generates a lot of spaghetti and it's a clear product that they're trying to just get you into this instant Dev products area so I think it's good for a lot of people that are very beginner because you could type in like make Spotify clone and it'll go and it'll do that and it'll generate a lot of stuff it reminds me um more of like repet and their agents where they're trying to make the full stack thing and it deals with the package dependencies and I think that's great but I think you lose a lot of control I think that you can't do as much tweaking as you like like something about running the ID in the browser to me just doesn't jive well and when I tried to get all the stuff out of the blitz or bolt when I was doing it it wasn't that great so I think I think this is just for a different type of developer I think it's interesting something to keep an eye on for sure but I don't necessarily love it and um that's all I have to say about that so the next thing I want to cover is just warp when it comes to terminals if you're using the terminal every day a lot of people don't like that you have to have a login for one but I don't pay a time for this and it's really helpful because I can use natural language so if I just type in command I then I can start typing things in natural language now this is really helpful if you have to do any sort of like maybe like a systemwide find where you're like trying to do uh telescope search across or you're trying to write a bash script or you're trying to do anything like oh I want to how do I run a migration in superbase I haven't used superbase in a while it'll run a search for superbase docs most of the time it won't even have to do that it'll just give you what you're trying to do but I can just start typing and it'll go and do that and it'll bring up the assistant I can also attach a context so it has attached context and I don't have to do command I all the time like if I just start typing and it's clear that this is not a shell command then it will automatically kind of flip on that little toggle there and next up this is pretty interesting so anthropics strengthens the US intelligence strong opener so they're partnering with paler now paler is the the big boy in the room when it comes to private defense and intelligence and so this is a big deal so what they're saying is that the makers of the claw chatbot they partnered with the big data analytics firm paler as well as AWS so they're giving us intelligence Defense Agency access to its models to boost operational efficiency across departments and help them make better and more informed decisions that's awesome that's great couldn't be better timing in my opinion for this to be happening and I found it's interesting just from a geopolitical standpoint you know we just have the election and a lot of things are shifting and I think that this will affect the way that we pick the tools we use as developers because you have one side which is the office that just came into power with Trump and you have him partnering basically he's got Elon in his family photos so I think that you know you had Sam Alman going to Washington about a year ago and it was trying to put a bunch of laws in place and trying to basically get legal capture and his backers were really big about that you have vode from kosla Ventures venod kosla talking about oh how would you deal with uh the Manhattan Project if we open source that that would be horrible and he's comparing nuclear weapons from World War I to an llm and I think that that's just fear-mongering so that they have defensibility it's a moat like there's no bigger Moe than a legal Moe because if you get it done and you have uh legislative capture around your industry I mean that's just a crazy amount of power and influ so something like this happening and something like this recent election really does kind of change things on what we will be using long term for Frontier models because Elon just unlocked you know he has 14 departments that have been holding back his companies so you know famously he got a rocket built and ready to launch faster than the US Department of whoever it was one of them could approve him and now with him putting $130 million to work through the Trump campaign and you know being part of the family photos and helping with that it's the biggest return on investment possible because he's not going to have any legislative issues he'll be able to move quickly I don't think he's going to do any sort of capture either and I think this is a big sign why was it anthropic over open AI how' they get that contract well I think it's because paler is very open open to open source they're open to working with people that don't have bad intention they want to work with the best we know that anthropic is the best right now when it comes to coding and code generation and big data I would assume whereas open AI maybe is lagging a little bit so I think that's interesting I think that we're going to have to keep an eye on xai and what they're doing because they stood up a larger data center than anybody in 17 days and there's a clip of Jensen Wang however you say his last name from Nvidia talking about how ridiculous it is that Elon was able to get together with with Larry Ellison and Michael Dell to put together this basically this super compute infrastructure and they did in 17 days and usually that takes years so I would be keeping an eye on them too that's something that you can do as a developer is you can get you know I think they're giving like $25 in credits out for the XI stuff it's not you know the 300 from Google but I haven't personally used the model yet but they're saying by December that it's going to be at the same level as GPT 40 which is huge in terms of timeline because it took years and years and years for open AI to get there it's taken just about a year for xai to get there so that's on there um you know one of the questions they had is like is Amazon holding anthropic to Ransom who knows not sure but yeah and then last I want to cover just some stuff on the agents so when it comes to agents let's open up this project agents are basically individual files that can do a task really well and I think that it can get a little confusing because some people try to stuff too much stuff into they try to stuff too much stuff they try to give an agent too much responsibility whereas multi- gentic workflows are much more effective and I want to walk through one that I did for a real business use case that I made over the weekend so the use case was I was doing some work I was doing an audit on a friend's application he needs some Dev help and I was there to help and it took a lot of time took you know 68 hours to put together this audit this this whole like report of like where is you know this the state of your product that involve going through all the screens and screenshotting them and then putting them into Whimsical and writing the notes about him and then doing some SEO stuff running Lighthouse running um you know the detailed extension to figure it out checking for errors actually running the project and afterwards I was like that just took too long there's there got to be a better way I can make an asent for this and that's where this idea came in to play with with this site auditor and so what it does from a high level is it uses a bunch of different tools to go out and do everything that I did manually but in like a minute and actually do more than what I did so from a high level it's doing visual testing where it's using the ability to basically capture screenshots of every single view that is available that's not behind o and you can do this including o but by the way I didn't have support for that cuz this is just a Saturday project but it does the visual testing so it's going and it's snapping photos of every single view for full screen page widths page Heights on desktop tablet and phone and that's important and then it's going and it's actually doing a performance analysis so it's running page speed Insight stuff how fast things load um it's also doing behind the scenes it's doing SEO audits so you can get domain Authority you can do page Authority you can do access ability evaluations so that's something that people really don't think about a lot but Google dings you on it and it makes the HTML better it's semantic um so things like the the screen reader text that no one thinks about it does security scanning it also does SEO optimization so the whole thing here was this agent basically has a bunch of the different uh files in it and it generates a report at the end so it's a CLI and in the CLI I can basically select what are the things that I want this to do and I can hit the space bar for all of them so if I don't want the screenshots or I don't want one piece then I can unselect it and then the argument that you pass is just URL so I could go and if my friend was like hey my site needs some help what's wrong with it or if I was running a Consulting agency I could do a list of these things I could do an array of URLs and have it generate reports and have it actually you know the output is not only a giant glob of Jason that we send off to anthropic with a prompt for both the report of customer but also a report of what the developer would do to make it better so it's doing all the steps and it's just done by saying all right well we're going to use the lighthouse and the Chrome launcher and these are the default categories we have performance accessibility best practices and SEO it has the different uh like screen that we want and so this would be an example of one agent where it's going and it's doing that and it's creating this summary at the end same thing with the audit progress so this is just something that's nice with a CLI that I added um this is all just written in JavaScript because typescript was being a pain and again Saturday project didn't feel like fighting with types but this is going through and it's saying okay well every five minutes or every however many seconds we're going to be doing some sort of update so I can see like how many pages does this app have like is it you know it's basically for the person that's using it you have the report generator which is using axios and it's generating both the dev report as well as a client report and summarizing the audit data so the output that you get from all these individual tools is basically like I said a gigantic raw Json and then that gets passed in we do some averages on it so if you have a lighthouse score for a th000 Pages you know you don't want to have all those out and then it gets sent off to Claud with a specific prompt that says hey this is what we need to do can you figure it out we're chunking things too so if you have a really long list you want to make it smaller uh you have the screen shot capture which I mentioned you have the framework detector so this was uh what I did was I actually had a prompt that I used to reverse engineer walizer walizer is I'll just show it's an extension for Chrome that shows you the Frameworks and all the tools that people are using so I was like how does this work and I basically fed a prompt using Zed saying hey look here's the API here's the landing page analyze the landing page figure out what this thing is doing what the customer value is and then analyze the API and the stack using the framework agent and figure out what it does and recreate it so that was really cool um and then yeah so you have the framework detector HTML validator so there's a lot of stuff that I didn't even know about like analyzing the semantic pieces like are you missing things it's kind of like running a linter on a server but you don't have access to it so it's running a linter on public bundle and then you have security Checker so this is using SSL is it you know what's your security policy um and then you know here's all the utilities that we have for the different options so I just wanted to show that off a little bit because it was really helpful for me and it I think it gives people an idea of what these agents can be used for one's headless so it's you know it doesn't have a UI but then you'd have you know a folder for each uh URL that you pass and then the test so you can take a look at it before and see if it was good or not and then you know you have the giant file that's actually doing the work so that is one of the agents and I think it's really cool and so that's the video for today I hope you guys found this helpful there was a lot of different stuff if you guys have anything that you want me to cover one thing that I made a video about already this week was the new model that Claude came out with haou I didn't find it very helpful I think it's great that there's a new model out but I didn't find it helpful and um I think that you know the ones right now to be keeping your eyes on are obviously everyone's waiting for 01 to drop we've have the preview but I'm still using that as Chain of Thought for architecting and figuring out plans so that one's something to keep an eye on for 01 and then anything that drops from xai or Claud moving forward is definitely something to pay attention to these are weekly videos where I do stuff like this and I sit down and it's going to be every Friday so if you like the video make sure you like it do do your uh good deed for the day and subscribe and I'll see you in the next one