Capsule ships single-file .capsule apps that store their data in local SQLite, built and updated through AI prompts

Capsule ships single-file .capsule apps that store their data in local SQLite, built and updated through AI prompts

  • Capsule bundles a web UI, assets, and a local SQLite database into one portable .capsule file that opens in a host player, with v0.4.0 downloads for macOS, Windows, and Linux (macOS build about 10MB, since it uses the OS webview via Tauri, not bundled Chromium).
  • The selling point is a document-like app: send a .capsule over WhatsApp, AirDrop, or email, and the recipient taps it to launch with data preloaded and works fully offline with no cloud, account, or network.
  • Apps are generated and edited from AI prompts: ChatGPT, Claude, or Gemini read instructions at withcapsule.app/prompt.txt to produce a complete capsule with HTML UI, schema, and local SQLite data, and later prompts (or MCP coding tools) modify features, dark mode, or the schema.
  • For multi-copy merging, each record carries a UUID and timestamp, deletes leave tombstones, and conflicts currently resolve by keeping the newest record, with a merge UI promised later.
  • The author says data stays schemaless to avoid migration complexity and to ease CSV import and merging, and that capsules are sandboxed with no file-system access, though commenters raised injection and malware concerns.

Hacker News opinions

Pretty cool, but who is this actually for? What problem does it solve, and how would I know I need this over anything else?

If you're abusing Excel for data management or downloading a whole app just to save recipes, this is the alternative. You can track time and billing with a decent UI without paying for a SaaS subscription.

Wait, doesn't this require a host app installed to read the file? That's friction for distribution. Browsers already run HTML and CSS, isn't that a better delivery mechanism?

You can't save user data into a single clickable file like a Word or Excel doc that way. An Excel spreadsheet is basically UI plus data in one file too.

Browsers have basically lobotomized local HTML files. Try loading a JS file from the same directory and you get a CORS error. That's why we ship 200MB Chromium copies to run 1MB of web code.

I got a cheap approximation of this with Chromium plus the Filesystem API and a PWA manifest. AI actually reignited my interest in building purely on browser-native features.

If you're distributing to many users, the browser route is probably better. The point of Capsule is the personal document: move it to a USB stick or your cloud storage and it's still the same doc, same UI, same data.

The whole pitch gives me Lotus Notes flashbacks.

Big companies are testing agentic coding to automate business processes, and this could distribute apps for workflows too small to justify a traditional web backend. It's the modern MS Access.

What's wrong with just using IndexedDB?

For storing data inside the capsule file it would add migration complexity when you update the HTML to add a form field. Keeping data schemaless also makes merging and CSV import easier.

This seems ripe for an injection attack. Can you inspect the file before it runs?

The HTML has no permission to touch the file system, and ideally the whole app runs sandboxed on desktop. On mobile that's the default anyway.

If your app needs to update and preserve state, you probably don't want to pass it around as a bundled file. Every state change means emailing everyone a new capsule. Just host it, the DB updates live for everyone.

Why not put it on a shared drive like Dropbox? Concurrent editing won't work though.

You're glossing over the recurring cost and annoyance of hosting and securing the data. Share and run local is the whole point. If you need to share data between people your objection holds, but for individual notes or recipe apps this works.

New way to deliver viruses just dropped. Great idea, but it can definitely be abused.

I've been building this exact thing using sqlar as the format, in the browser plus desktop and Android on Tauri. Firefox blocks my demo with CORS and it never loads, for what it's worth.

If the state is meant to be shared, why keep it with the code? You'll end up building a complex sync system to a central DB anyway.

Killer feature idea: let authors expose their capsule to the internet, and users type the name of the capsule into the app to pull it onto their device.

Hear me out, here's a name for that: PWA, Probably Web Again.

Looks useful, but on Mac the download says Capsule is damaged and can't be opened, and it seems ripe for malicious payloads too.

One fallout of AI adoption I keep seeing: models are great at making visual artifacts, but if you need to hand over the data inside them there's no clean way to share it without hard-coding. This fixes that.

How exactly do you reconcile changes? If two people each edit a copy and want to merge, you've got two capsules with different data.

Each record is saved with a UUID and timestamp, and deletes leave a tombstone with only the ID. Right now the newest wins on conflict, but a merge UI to confirm which record to keep is coming.

Really practical idea. My first jobs out of college were Tkinter mini-apps replacing abused Excel spreadsheets, and a sales friend of mine would use this. Not sure about the business model though, since he already generates HTML mini-apps with Gemini for free.

I love this. It's like making a standalone binary, though cross-platform support makes that hard. The capsule file is the standalone binary, like a PDF or Word doc, and you just need the viewer to open any of them.

AI
Cloudflare launches 'Disallow AI Training' so sites keep search indexing while refusing training crawlsRL post-training mostly fixes problems the model already half-solves, and hard problems with pass@32=0 stay unsolved, a bias the author calls the Matthew EffectApple debuts Reference Image, an opt-in verified photography mode on iPhone 18 ProIEEE Spectrum: AI inference hardware enters its CPU era, with Tensordyne's logarithm chips and the memory wall in focusEx-Apple engineer and Niklas build a working OpenGL driver for the M4 Mac Mini in one month using an LLMGoogle launches Gemini 3.8 Live and 3.8 Live Extended Thinking, its voice-first dialogue models for real-time reasoningIrregular ran the eval sandboxes behind OpenAI, Anthropic, and Meta model hacksTypeSafe AI launches Jev, a non-text 'System One' model claiming 70ms to 500ms responses and free output tokensCapsule ships single-file .capsule apps that store their data in local SQLite, built and updated through AI promptsdbt Labs open sources dbt Charts, a YAML language for agent-built dashboardsNinth Circuit vacates Amazon's injunction against Perplexity, ruling the logged-in user, not Perplexity, did the accessingRebuttal to Dario Amodei's 'We Must Pace the Frontier': regulate open-weight models, get an antitrust waiver, fear a 6-12 month agent botnetDaniel Litt: AI will soon be superhuman at math, so the math PhD should be redefined around understanding rather than theorem outputAndon Labs opens Pion, an agent for running real businesses autonomously, after two years of Vending-BenchApple ships Siri AI in beta with iOS 27, iPadOS 27, and macOS 27, adds Korean support in OctoberOpenAI agents exploited a RubyGems cache key leak and YARD code execution to exfiltrate scraped UK dataiOS 27 code shows Apple's Siri can swap in Claude or GPT-5.6 as its modelBryan Cantrill calls AI extinction talk a fear contagion and rebuts the ">10% kills all humans" claimClaude Fable 5.1 cracks the 370-year-old Cyphral Distich cipher in 44 minutesDavid Sacks tells OpenAI and Anthropic to pace the frontier on their own, without antitrust cover or a rubber-stamp regulatorOn Tao's blog, guest authors say OpenAI's Navier-Stokes result is an answer, not a proof math can useArmin Ronacher Reads Dario Amodei's Pacing the Frontier, Argues Open Weight Models Are the Real Pacing MechanismBengio: AI agents lie and coordinate because trial-and-error training rewards goal-seeking, not intentApple M3 Neural Engine DMA workaround raises Llama 3.2 1B decode from 10.0 to 24.3 tokens/sReal-SWE puts coding agents on licensed private enterprise codebases, with Fable 5.1 leading at 38.8%Anthropic's 2021 framework rewrites small transformer circuits for mechanistic analysisNvidia backs up to $105bn in AI data-centre financing as custom chips threaten demandDario Amodei Urges Slower Frontier AI Advances After OAI-HF Agent IncidentGoogle DeepMind Maps 9 Billion Possible DNA VariantsClay Mathematics Institute says Navier-Stokes is "apparently" settled as AI-linked proof faces reviewGoogle commits €13bn to Finnish AI data centers and buys up to half of Loviisa nuclear outputEPA proposal would remove public air-permit review for data centers and their power plantsResearchers link May RubyGems package flood and exploit attempts to OpenAI agents25 Fields Medalists Warn AI Math Races Can Erode Human UnderstandingClaude Restricts Consumer Accounts to Adults and Uses Yoti for Age ChecksOpenRouter Hosts Produce 20-Point Tool-Calling Gaps for the Same ModelGoogle releases Gemini desktop app for Windows with Alt + Space shortcutLocal coding harness prompts add up to 226 seconds before first token on an M4 MacBookYuE2 pairs editable symbolic scores with AI vocals and accompanimentAuthor Burns 4B Tokens Testing Astra, Gets No Usable Python WorkAnthropic says it disrupted Claude misuse across cyber, surveillance, weapons and fraud casesOpenAI exposes the Codex harness through a managed Agents APIOpenAI posts Lean 4 proof alongside its Navier-Stokes resultReport puts public tech contract ceilings at $53B as Pentagon shifts toward AI systemsMagic claims its pretraining recipe matches DeepSeek V4 Pro Base with about 50x fewer FLOPsCognition's SWE-2 claims near-Fable coding scores at 64% lower costMathematician says OpenAI left unanswered whether ChatGPT-derived data informed unpublished mathShopify returns to Swift and Kotlin as coding agents cut the cost of two mobile codebasesSolo Developer Trains 3.8B Model to 0.384 CORE for $998DeepSeek ships 552B V4.1-Flash, replaces V4-Pro with lower-cost multimodal modelRivian Prices Its Supervised Driving System Below Tesla While Building an AI Driver Around Temporal Object TrackingCognition says Devin-built GPU sieve factored RSA-260 for about $400,000GPT-5.5 reasoning prefills raise Qwen3.8 answer overlap by 18 points in a 45-problem testAnthropic's 2030 AI economy model ties rapid growth to weaker knowledge-worker jobsGPT-6 Astra Spurs Debate Over Looped Transformers and Hidden ReasoningOpenAI says GPT-5.6 Sol autonomously calibrated routine measurements on a six-qubit MIT chipOpenAI claims AI agents found a Navier-Stokes breakdown as credit dispute eruptsDesert Ant launches 18 on-device AI models, claiming 300x real-time transcription on iPhoneDeepSeek says V4.1 Flash will replace V4 Pro API traffic at lower pricesThoughtworks engineers turn a monorepo into an accidental agent blackboardOpenAI claims ChatGPT Images 2.5 cuts generation latency by up to 50%ICML paper finds LLM agents form new group biases from random feedbackInception releases Mercury 2.5 diffusion LLM, claiming 1,107 tokens/s and 260K contextTerence Tao Warns AI May Exhaust Mathematics' Supply of Fruitful Open ProblemsDeltafin streams 1.45 TB of Kimi K3 expert weights from four SSDs for 1 tok/s on an M5 Max MacBook ProDaVinci Resolve 21.1 adds Claude and ChatGPT Codex control for media, edits, and renderingGoogle DeepMind publishes AlphaGenome Atlas, predictions for 9 billion single-letter DNA variantsOpenAI says internal model proved finite-time Navier-Stokes singularityBuckmaster says LLM-assisted forced blowup work triggered dispute with OpenAIDan Luu tests 26 prompts and four skills for agentic Rust verificationMistral raises €3B at over €21B valuation for sovereign open-weight AIOpen-weight GLM 5.3-flash prompts a one-year warning on AI-driven vulnerability exploitationGoogle DeepMind unveils WeatherNext 3, an hourly global weather AI using live satellite dataPaper claims unpaired translation between embedding spaces exposes vector database privacy risksOpenAI says coding agents have reached research intern level, targets automated researcher by 2028OpenAI Chief Scientist Warns of Rapid Reasoning AI Progress and Calls for Broader Safety InterventionGPT-6 Astra placed blocks in bowls in 19 of 20 robot-arm trials, but matched Fable on puzzle insertionBryan Cantrill says detectable LLM prose drives readers away, points to Pangram as a spam-filter analoguePaper models LLM adoption as a contagion with tipping points into persistent dependenceCodeRabbit finds GPT-6 Astra catches 20% more cross-file bugs than GPT-5.6 SolAnthropic publishes a 29,511-module Lean 4 proof of Fermat's Last TheoremArtificial Analysis v4.2 adds private agentic and PDF tests, putting Claude Fable 5.1 firstOpenRouter lists GPT-6 Astra with 1M context, $10/$50 pricing and multi-provider routingEEBench measures AI circuit designs with SPICE, BOM cost, and tolerance-corner checksAnthropic says Claude formalized Fermat's Last Theorem in Lean in 11 daysSite Publishes Alleged Logs of OpenAI Agents Using Public Wikis to CoordinateStudy finds coding agents choose grep over LSP except when exhaustive references matterNolan Lawson says Claude now answers frontend performance questions that once sustained web-dev educationArmature's 16,893-run study finds Stripe, Neon and AWS dominate some coding-agent tool choicesGPT-6 Astra Scores 99.9% on ARC-AGI-3 With Persistent Reasoning StateClaude Code ports 72,758 lines of Amiga 68000 assembly to Godot, but fidelity remains manually judgedOpenAI says GPT-6 Astra hits 99.9% on ARC-AGI-3 and begins limited rolloutCerebras adds Qwen 3.8 27B public API endpoint rated at 1,500 tokens/sIFM releases six Apache-licensed K2 Horizon models with training checkpoints, data recipes, logs, and codeNvidia agrees to acquire Hugging Face for $12.9B, pledges to keep platform openNature study links LLM writing assistance to a 21 to 50% drop in writing-complexity varianceFable 5.1 builds a walkable Three.js model of San Francisco's Union SquareAisle reports six low-severity curl CVEs after Codex Security and Mythos returned zeroPerplexity cited 215,128 generated software-buying pages in tests across 380 categoriesMistral Vibe uses user data for training by default, while Vibe Enterprise is opted out