🛠️ All DevTools

[Other] Agent-talk: Enabling coding agents to work together

[API/SDK] Show HN: Open-source AI app builder you can embed into your own SaaS We just open sourced a first class our AI web app builder.<p>Instead of using another hosted AI coding platform, you can fork this project and build your own AI app builder, fully customized and running under your own brand.<p>It includes:<p>Next.js + TypeScript AI chat with streaming Artifact generation File explorer Code editor Live preview Databases Sandboxes to be used by AI agents Versions Responsive production-ready UI And a lot more...<p>The only required dependency is the Totalum API, which exposes the AI generation engine through a simple REST API. You can replace or extend the backend however you want.<p>It&#x27;s designed for developers, SaaS companies and agencies that want to:<p>Build their own AI app builder Add AI app generation to an existing product Create a white label AI builder Customize every part of the UI and workflow Self host the frontend<p>We use the same frontend in production, and decided to open source it so others don&#x27;t have to start from scratch.<p>Repository: <a href="https:&#x2F;&#x2F;github.com&#x2F;totalumlabs&#x2F;ai-app-builder-open" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;totalumlabs&#x2F;ai-app-builder-open</a><p>I&#x27;d love feedback, feature requests, or ideas from the HN community.

[Testing] Show HN: Sentinel – open-source QA agent that reads your code before it clicks

[Other] Show HN: Docs.dev Your Own Hosted Docs Platform in Minutes Hey we are launching Docs.dev, a full feature docs product that lets you host your developer docs on Cloudflare in a couple minutes.<p>Why this exists:<p>We don’t think every early stage startup should need to pay $300-$500 a month for beautiful docs, in fact every startup should just have beautiful human and agent ready docs ready to go. Additionally the workflow for docs is a lot different now in that Claude Code or Codex will generate 60-80% of the code, so why not have them update your docs as well. After they generate the docs, the missing piece is there will be some sentences, sections, and codeblocks maybe you want to eyeball or edit. The Docs.dev docs allows you to do it right in place in that your docs powered by Docs.dev, is the place that both your readers can learn and read about your tool but also the direct place you can edit your docs.<p>If you want to check out how it works see the demo at <a href="https:&#x2F;&#x2F;docs.dev" rel="nofollow">https:&#x2F;&#x2F;docs.dev</a>.<p>Some features:<p>- Ask AI using your own Cloudflare AI - Docs are editable and publishable in place exactly how users see the docs - AI generation tools built into the editor(again using your own Cloudflare) - Full markdown support for agents, every docs page can be served as MD - Docs can be served as MCP - Pretext.js editor allows images to be inline with the text<p>Built upon other cool technology:<p>- Fumadocs - Pretext - Cloudflare primitives - Other cool libraries<p>This is a side project from us at Keyboard.dev, so right now we want to support making this product better but we are looking at this more as a gift to the community. Essentially before our pivot we had the domain for docs.dev so wanted to build something cool in the docs space.

[Other] Guide to data tools landscape for developers

[Other] We're Building Postgres in Rust. Using the LLVM of Databases

[Other] Show HN: Trying to fix the web scraping industry's benchmark problem We&#x27;ve been trying to evaluate web scraping companies, but when you look at their benchmarks, you can&#x27;t verify anything, and they mostly exist to prove the company is successful. They put somewhere between 98% and 100% because they pick their own urls, define success their own way, and don&#x27;t publish the harness. We also saw companies like scrapfly astroturf websites like scrapeway and call them independent.<p>So, we built an open source benchmark that we want to represent the frontier of web data. We&#x27;re trying to look across all major anti-bot providers and industries, to build a comprehensive hard-target test set. It&#x27;s all in the repo, you can run it, and we pledge to maintain it monthly.<p>Curious what people think we could do to improve it? Would you change any of the conditions, add sites, etc. Our target list and pass criteria are something we&#x27;re trying to improve, so would love to get feedback from the community on how we can actually provide something useful for people to test against!

[Other] Show HN: Forall – An AI coding agent that generates machine-checkable proofs We&#x27;ve been working on an open-source coding agent that generates code alongside machine-checkable proofs. We&#x27;d love feedback from the HN community, especially from people interested in formal verification, Lean, Dafny, or AI coding agents. Currently, only 3 langauges can be verified.

[Other] Show HN: Firefox in WebAssembly This is the entire Firefox browser rendering to a &lt;canvas&gt; element. Gecko, all UI components, and the Spidermonkey JS engine are all compiled and running in WebAssembly.<p>Here are a few things you might find interesting:<p>- This is fully end to end encrypted! We use the WISP protocol for TCP-over-websockets.<p>- There is a novel WASM-&gt;JS JIT for experimental site speedup<p>- This port cost over 25k in opus&#x2F;fable tokens for debugging and JIT research<p>This was just a fun experiment to push the boundaries of WebAssembly. For a more usable &quot;browser in browser&quot; experience, we also built <a href="https:&#x2F;&#x2F;github.com&#x2F;HeyPuter&#x2F;browser.js" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;HeyPuter&#x2F;browser.js</a> that eats a bit less RAM.

[Other] Grok Build is open source

[Other] Brainless: Shadcn components that look like Claude Code, Codex and Grok

[Other] Open-source memory for coding agents, synced over SSH

[Other] Show HN: misa77 - a codec that decodes 2x faster than LZ4 (at better ratios) I&#x27;ve spent the last few months working on this codec.<p>It has the following characteristics:<p><pre><code> - SOTA decompression throughput in its ratio class - Decent ratios (comparable to LZ4 at high effort levels) - Slow compression </code></pre> Most of the gains can be attributed to reducing branches and making decompression very friendly to out-of-order cores, by using a smart format.<p>Results on the tarred Silesia corpus on Intel x86-64 follow:<p><pre><code> codec decode ratio encode misa77 -0 5219 MB&#x2F;s 42.64% 54.5 MB&#x2F;s misa77 -1 4274 MB&#x2F;s 39.65% 51.2 MB&#x2F;s lz4 2505 MB&#x2F;s 47.59% 371 MB&#x2F;s lz4hc -12 2531 MB&#x2F;s 36.45% 7.31 MB&#x2F;s</code></pre>

[API/SDK] Launch HN: Coasty (YC S26) – An API for computer-use agents Hey HN, we’re Nitish and Prateek, the founders of Coasty (<a href="https:&#x2F;&#x2F;coasty.ai&#x2F;computer-use">https:&#x2F;&#x2F;coasty.ai&#x2F;computer-use</a>). We’re building computer-use agents that can complete workflows inside legacy desktop software and web applications without usable APIs.<p>Developers send Coasty a natural-language task either through our consumer app or through our API, select a machine or browser environment, and any relevant credentials or files. The agent then operates the interface through screenshots, mouse, and keyboard input, verifies the result, and returns a structured run record with screenshots, actions, outputs, and errors.<p>Here is a raw demo of an agent completing a workflow in a legacy application(It’s a mockup): <a href="https:&#x2F;&#x2F;drive.google.com&#x2F;file&#x2F;d&#x2F;1ZghU_3vsAYhHVz1bsvE0pkvZYk7OUnb1&#x2F;view?usp=sharing" rel="nofollow">https:&#x2F;&#x2F;drive.google.com&#x2F;file&#x2F;d&#x2F;1ZghU_3vsAYhHVz1bsvE0pkvZYk7...</a><p>A lot of important software is still difficult to automate. Healthcare teams submit prior authorizations through payer portals, accounting teams enter data into desktop applications, and operations teams move information between internal systems, spreadsheets, and remote desktops. Many of these applications have no API, incomplete APIs, or integrations that take months to build.<p>The usual alternative is RPA, record a sequence of clicks and replay it. That works when the interface and workflow are predictable, but it often breaks when a button moves, a pop-up appears, a page loads slowly, or the application enters an unexpected state.<p>Coasty takes a different approach. The agent observes the current screen, decides what action to take, executes it, and then observes the resulting state before continuing. It does not require DOM access, an accessibility tree, selectors, or an application-specific integration, so the same API can operate browsers, remote desktops, and older Windows applications.<p>A simplified request looks roughly like this:<p><pre><code> run = coasty.runs.create( environment=&quot;vm_123&quot;, task=&quot;&quot;&quot; Open the patient record in the billing portal. Enter the attached authorization data. Do not submit if the member ID or procedure code does not match. Return the confirmation number. &quot;&quot;&quot;, files=[&quot;authorization.pdf&quot;], approval_required=[&quot;final_submission&quot;] ) </code></pre> The response includes the final status, extracted outputs, a replay URL, and a timestamped event log:<p><pre><code> { &quot;status&quot;: &quot;completed&quot;, &quot;output&quot;: { &quot;confirmation_number&quot;: &quot;PA-184392&quot; }, &quot;replay_url&quot;: &quot;...&quot;, &quot;events&quot;: [ { &quot;type&quot;: &quot;verification&quot;, &quot;field&quot;: &quot;member_id&quot;, &quot;result&quot;: &quot;matched&quot; } ] } </code></pre> The API can also pause a run for human approval, retry from a checkpoint, or return control to the developer when it encounters a condition the workflow did not anticipate.<p>We started working on this last summer, because we saw that models were getting better at vision but kept seeing a gap between computer-use demos and the reliability needed for production workflows. Getting an agent to complete a task once is fairly straightforward. Getting it to repeat that task, recover from unexpected states, avoid silently entering incorrect data, and produce evidence of what it did is much harder.<p>We built several layers around the underlying computer-use model. The system tracks the expected state of the workflow, detects when the application has diverged from that state, and can re-plan instead of continuing blindly. Developers can define invariants such as “the patient name must match the source document” or “never submit without approval,” and the agent checks those conditions during the run.<p>Each run happens in an isolated virtual machine. We expose APIs for provisioning environments, uploading files, starting tasks, streaming events, inserting human approvals, and retrieving the full replay and audit trail. Environments can be kept alive across runs when the application has a long login flow or persistent local state.<p>One problem we are still working through is the tradeoff between speed and reliability. The agent can move faster by taking fewer observations and verification steps, but that becomes risky in workflows involving patient records, payments, or regulatory filings. We currently bias toward slower execution with more checks and let developers configure approval points and verification policies.<p>We are initially working with healthcare operations teams because their workflows combine many of the hardest conditions: payer portals, EHRs, PDFs, spreadsheets, remote desktops, and actions where quiet mistakes are expensive. We also expose the same infrastructure through the developer API for teams building their own agents and vertical automation products.<p>We currently charge based on agent runtime and workflow volume, with separate pricing for dedicated environments and enterprise deployments.<p>We’d especially appreciate feedback from people who have built and&#x2F;or used browser agents, RPA systems, desktop automation, or agent infrastructure. We’re curious which parts of the API you would want direct control over, where you would prefer higher-level abstractions, and which failure modes have been hardest in your own automation systems.<p>If you&#x27;ve hit weird failure modes automating software like this, we want to hear about them. We&#x27;ll be here all day answering questions and taking notes!

[Database] Show HN: SirixDB 1.0 Beta – Git-Like Versioning, Diffs, Time-Travel Queries Hi HN! I&#x27;ve posted SirixDB here before, back in 2019 (<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=19834681">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=19834681</a>) and again in 2023 (<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=38252963">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=38252963</a>).<p>The core idea behind SirixDB is, that history is a first-class citizen. Every commit stores a lightweight, queryable revision. You can query any point in time, even individual nodes (for instance JSON values), diff arbitrary revisions, and efficiently track how data evolved without replaying events.<p>Unlike traditional event stores, historical states do not need to be reconstructed by replaying events nor do we have to think about projections. Revisions are directly queryable.<p>A simple example:<p>Jan 1: Record &quot;Price = $100, valid from Jan 1&quot;. Stored on Jan 1 (transaction time).<p>Jan 20: Discover price was actually $95 on Jan 1. Commit correction.<p>After correction, you can ask across both axes:<p>- &quot;What did we THINK the price was on Jan 16?&quot; -&gt; $100 (Transaction time)<p>- &quot;What WAS the price on Jan 1?&quot; -&gt; $95 (Valid time)<p>I&#x27;ve worked on this in my spare time since 2013, following its academic precursor (Idefix&#x2F;Treetank) at the University of Konstanz. The architecture relies on an append-only physical log and a persistent copy-on-write page trie.<p>A high level view of the architecture:<p>Physical Log (append-only, sequential writes)<p><pre><code> ┌────────────────────────────────────────────────────────────────────────┐ │ [R1:Root] [R1:P1] [R1:P2] [R2:Root] [R2:P1&#x27;] [R3:Root] [R3:P2&#x27;] ... │ └────────────────────────────────────────────────────────────────────────┘ t=0 t=1 t=2 t=3 t=4 t=5 t=6 → time </code></pre> Each revision is indexed, and unchanged pages are shared:<p><pre><code> [Rev 1] [Rev 2] [Rev 3] │ │ │ ▼ ▼ ▼ [Root₁] [Root₂] [Root₃] │ │ │ │ │ │ │ └─────────┐ │ └────────┐ │ └─────────┐ ▼ ▼ ▼ ▼ ▼ ▼ ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ │ P1 │ │ P2 │ │ P1&#x27; │ │ P2&#x27; │ └──────┘ └──────┘ └──────┘ └──────┘ Rev 1 Rev 1+2 Rev 2+3 Rev 3 (shared) (shared) </code></pre> Beneath the root pages sit node and secondary indexes, using a novel sliding-snapshot algorithm to balance read&#x2F;write performance. Everything is queryable using JSONiq via the Brackit compiler.<p>Back in 2019, and even in 2023, SirixDB was very slow due to GC pressure. Unlike most other document stores, SirixDB stores fine-grained nodes, and I came to realize that an on-heap (JVM) representation made up of lots of small objects simply didn&#x27;t make sense. I measured it with async-profiler — with some help from Andrei Pangin himself — and the result was that the poor throughput was due to the sheer amount of allocations which scaled almost linearly with the number of open transactions.<p>Working a full-time software engineering job, I lacked the energy for a massive spare-time rewrite. About a year ago, I started experimenting with AI. It turned out to be ideal for automating the tedious, repetitive parts of migrating the storage layer to Java&#x27;s Foreign Function &amp; Memory API, storing pages completely off-heap.<p>Looking further ahead, the append-only, immutable-page design maps naturally onto object storage like S3 and distributed logs like Kafka for a cloud version, and initial prototypes already exist. Maybe that becomes a commercial service one day, but for now, I&#x27;m just thrilled to see these core design principles finally proven out.There&#x27;s an interactive demo, documentation, and the code is on GitHub. I&#x27;d love feedback and am happy to answer questions!<p>kind regards<p>Johannes<p>[1] <a href="https:&#x2F;&#x2F;sirix.io" rel="nofollow">https:&#x2F;&#x2F;sirix.io</a> | <a href="https:&#x2F;&#x2F;github.com&#x2F;sirixdb&#x2F;sirix" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;sirixdb&#x2F;sirix</a><p>[2] <a href="https:&#x2F;&#x2F;sirix.io&#x2F;docs&#x2F;architecture.html" rel="nofollow">https:&#x2F;&#x2F;sirix.io&#x2F;docs&#x2F;architecture.html</a><p>[3] <a href="https:&#x2F;&#x2F;demo.sirix.io" rel="nofollow">https:&#x2F;&#x2F;demo.sirix.io</a><p>[4] <a href="https:&#x2F;&#x2F;sirix.io&#x2F;docs&#x2F;" rel="nofollow">https:&#x2F;&#x2F;sirix.io&#x2F;docs&#x2F;</a><p>[5] <a href="http:&#x2F;&#x2F;brackit.io" rel="nofollow">http:&#x2F;&#x2F;brackit.io</a>

[Other] Show HN: StyleSeed – a design-rules engine so AI agents stop building generic UI

[Other] Show HN: Leet Robotics: Learn robotics and ROS2 with hands-on courses Hi all,<p>I&#x27;ve just launched Leet Robotics: a platform to learn robotics hands-on, with a full ROS2 workspace that runs in the browser (Jazzy, Gazebo Harmonic, Foxglove, VS Code) - no install required. The platform also has room for sharing projects and simulation assets as it grows.<p>Our first course is live now: Intro to ROS2 (free to read). The course teaches skills ranging from building your first node to a capstone project of a robot touring a museum world, with every lesson runnable in the online workspace (free accounts get an hour of workspace time daily - enough to follow the course).<p>Would love feedback from this community: on the course, the workspace experience, and what courses to build next.

A coding agent for low-cost models

[Other] Show HN: Make senders work to get into your inbox Hi HN :) really excited to share this with you.<p>The one thing AI reliably does is generate noise. Half the tools I see launch are just machines for producing more noise across more channels. And people are starting to see this in the form of emails in their inboxes as spam filters are struggling.<p>There used to be a useful signal in email: the effort a sender put into customizing a message was a rough proxy for how relevant it actually was. AI killed that. Now it&#x27;s customized slop with the appearance of effort with none of the cost. It is painful that the open internet &#x2F; open channels have been abused like this.<p>Captchainbox applies the idea of proof-of-work to email. If a sender is willing to do a bit of work to reach you, the message is more likely to be worth your time and the sender more likely to be real. The work is a traditional captcha. You can also set a pay-to-deliver amount if you want more friction. The proceeds of the delivery payment after transaction costs go to the Internet Archive and the EFF. The tool currently works by authing with your Gmail or Outlook and during launch time I make this completely free as a lifetime deal (with optional payment if you wanna support).<p>How it works: Captchainbox builds a whitelist automatically from the metadata of your past correspondence. If you&#x27;ve emailed an individual address, that sender can reach you. If you talk to several people at the same domain, we whitelist the whole domain. If one transactional-looking sender has sent you more than 10 emails, we treat it as a transactional domain and let it through. This whitelist is for you to change whenever you want. It continues to build organically as you converse with more addresses.<p>Incoming mail is checked against that whitelist. Senders already on it land in your inbox as normal. Anyone else gets archived (never deleted) and is sent a challenge. This can be the captcha or the payment link. Once they solve it, their email is pulled out of the archive and put back into your inbox.<p>if you want to see what this looks like from a sender&#x27;s point of view, send me an email here: doerpfelix15@gmail.com<p>The service only ever reads metadata, never message content. And since nothing is ever deleted, you can&#x27;t lose a message. There is a legitimate risk &#x2F; downside: if you sign up to a new service, these emails also land in the archive. Since we do not process the content, a first-time sender who can&#x27;t solve the challenge (say an automated activation email) will sit in your archive until you spot it.<p>Happy to answer anything! :)

Jurassic Park computers in excruciating detail