scripts - Moozonian Search

vyvoice.com• May 16, 2026• 1 min read

Show HN: Vyvoice: Privacy-first, cross-platform, offline voice transcription appHey Hacker News,vyvoice is a cross-platform, offline voice transcription app I started working on in December as a Windows user tired of every good dictation app being Mac-only.Beyond transcription, it has built in support for voice commands & gen-AI agents (work-in-progress). Windows, Linux, and macOS are currently supported. The primary product is free, agentic AI and other future premium features will be behind a subscription ($5/month).Underneath the hood vyvoice leverages ONNX to provide transcription via Parakeet and Whisper models locally on the device. I've developed an efficient VAD loop to ensure only segments of audio that contain speech are being processed for transcription. Valid audio segments are decoded in real time with end of utterance. Your audio and transcripts never leave your device.I would love feedback on all aspects from default settings to commands and agent functionality. I am actively working on the agent feature and plan to support skills, MCP, and more in

github.com• May 13, 2026• 1 min read

Show HN: SelfCertForge, manage root CAs and self-signed certs on macOS/WindowsHi, HN!I recently published SelfCertForge, an open source desktop app for creating and managing self-signed certificates on macOS and Windows:https://github.com/rbonestell/SelfCertForge/I know, I know... this isn't an earth-shattering invention. This came from a place of years of personal annoyance. I’ve been maintaining scripts that wrapped OpenSSL for local certificate workflows: create a long-lived root cert, add it to the trust store, then generate and sign child certificates for local services.The scripts work but the workflow was clunky. Used infrequently enough that it's easy to forget the flags, not much easier to explain to anyone than OpenSSL itself, and irritating to repeat across machines. Every time I'd share them with colleagues and friends there were always questions and feature requests, so I used Claude Code to amplify my UI/UX design skills (well, I can't amplify 0...) and turned the functionality into a cross-platform GUI.SelfCertForge can generate root CAs and add t

bing.com• May 12, 2026• 1 min read

Billions In Prescriptions Go Unfilled. This Startup Is Using AI To Fix That.Forus reached $1 billion valuation by dealing with the unglamorous admin work of processing scripts so patients can get the drugs they need. Nearly one third of Americans never fill the prescriptions ...

github.com• May 11, 2026• 1 min read

Show HN: adamsreview – better multi-agent PR reviews for Claude CodeI built adamsreview, a Claude Code plugin that runs deeper, multi-stage PR reviews using parallel sub-agents, validation passes, persistent JSON state, and optional ensemble review via Codex CLI and PR bot comments.On my own PRs, it has been catching dramatically more real bugs than Claude’s built-in /review, /ultrareview, CodeRabbit, Greptile, and Codex’s built-in review, while producing fewer false positives.adamsreview is six Claude Code slash commands packaged as a plugin: review, codex-review, add, promote, walkthrough, and fix. I modeled it after the built-in /review command and extended it meaningfully.You can clear context between review stages because state is stored in JSON artifacts on disk, with built-in scripts for keeping it updated.The walkthrough command uses Claude’s AskUserQuestion feature to walk you through uncertain findings or items needing human review one by one. Then, the fix command dispatches per-fix-group agents and re-reviews the work with Opus, reverting a

useslate.app• May 8, 2026• 1 min read

Show HN: Slate – agentic pre-production studio for solo YoutubersI built slate as a personal tool to centralize my strategy, research, scripting, thumbnails and shots in one place. Started showing it to other youtubers and that made me wonder if more people could have the same problem as me.Here's a quick walkthrough: https://youtu.be/7Ml5dVcFFdsI've been doing Youtube videos since I was a teenager. There's so many work before even recording: researching what's working in your niche, finding outliers worth studying, obsessing over titles and thumbnails that will earn you the click, writing scripts that actually hold viewers attention and somehow keeping your B-roll from turning into a mess.Big channels hire teams for this. Solo creators struggle juggling dozens of tabs, google docs and notion templates.I'd love to hear for feedback.

github.com• May 5, 2026• 1 min read

Show HN: Agent Historic Philosophical Persona Routing and PromptsI've been building this for a while. The core of it is I've assigned different tasks in a software engineering job to different philosophers. The prompts are written toward their thinking by "them" but the keywords are generic. I absolutely love using this prompt system. I share it every chance I get. I wrote a whitepaper about it. Maybe ya'll would be interested in it. I find my experience is better with my prompts than without.The only thing you might not like is my heavy handed treatment of logging output. I was seriously fed up with LLMs trying to run huge tests or scripts and then looking at just the last 20 characters. Everything is logged out to files and then the files are queried for the answers it wants.

allbsides.com• May 4, 2026• 1 min read

Show HN: I indexed 8,643 BSides talks across 227 chapters and 6 continentsHi HN,I'm Roland, and for the past few weeks, I've been building AllBSides — a directory of every BSides conference talk uploaded to YouTube. As of today, 8,643 talks from 5,927 speakers across 227 chapters in 68 countries. Combined runtime is 280 days. The transcripts come to about 60 million words.The archive came together in stages:1. Manually map every BSides chapter's YouTube channel 2. Pull every video and transcript from Supabase 3. Run each transcript through Haiku for tag extraction (tools, topics, difficulty, team, talk style, research method, and much more) 4. Run results through Sonnet for categorization and dedup 5. Final pass goes through Opus for verification 6. Do a manual verification - at one time, the pipeline showed over 16k AI suggestions for manual verification. Today, most are resolved.Total LLM cost so far: about €200. The whole pipeline is rebuildable from scratch.Each talk gets its own page with embedded video, full transcript, speakers, tags, and "related tal

longevity-tools.com• May 2, 2026• 1 min read

Show HN: I built Male Hormone Lab Interpreter that does what LLMs can'tHi,My name is Zsolt. I build clinical tools for labs and longevity clinics, and I release some of them as free public tools.The Male Hormone Lab Interpreter unique feature is that it visualizes the testosterone production and feedback pathway on HPG axis to identify the true root cause of hormonal imbalances. All hormones are interpreted based on optimal ranges. This is something LLMs still struggles with.The only difference between public and clinical ones is that the clinical suggest diagnosis and treatment options. But if there is any type of disfunction its obvious from the diagram and differentiating is also straigh forward.All the tools on the longevity-tools.com are 100% free and private. No email or any data collection, no third party scripts, no tracking scripts. Full privacy.Check out my other tools:## Thyroid Function Interpreter https://longevity-tools.com/thyroid-function-interpreterVisualizes hormones, calculated sensitivities, and secretory capacities on the HPT axis so

interfaze.ai• Apr 29, 2026• 1 min read

Show HN: A new benchmark for testing LLMs for deterministic outputsWhen building workflows that rely on LLMs, we commonly use structured output for programmatic use cases like converting an invoice into rows or meeting transcripts into tickets or even complex PDFs into database entries.The model may return the schema you want, but with hallucinated values like `invoice_date` being off by 2 months or the transcript array ordered wrongly. The JSON is valid, but the values are not.Structured output today is a big part of using LLMs, especially when building deterministic workflows.Current structured output benchmarks (e.g., JSONSchemaBench) only validate the pass rate for JSON schema and types, and not the actual values within the produced JSON.So we designed the Structured Output Benchmark (SOB) that fixes this by measuring both the JSON schema pass rate, types, and the value accuracy across all three modalities, text, image, and audio.For our test set, every record is paired with a JSON Schema and a ground-truth answer that was verified against the sou

bing.com• Apr 26, 2026• 1 min read

What Does the Editorial Board Do, Anyway?Scholarly authors often feel unsure about what happens behind the closed doors of academic publishers once their manuscripts are accepted for peer review. While most authors know about the peer-review ...

read.tube• Apr 26, 2026• 1 min read

Show HN: ReadTube – Turn YouTube subscriptions into a personal SubstackI built ReadTube so that I can read my YouTube subscriptions.YouTube has lots of high quality content. However, videos can be difficult to consume efficiently, especially those that are long and about series topics (e.g. general relativity, quantum physics). So I created this app to solve this need.There are already many existing YouTube AI transcription websites. However, ReadTube is one step further in that it periodically fetches the updates from the channels I subscribe to so I don't need to paste in the video URL each time. Also it can generate two different versions of summary: one short and one long, as well as a full length article. I can pick which length to consume according to the type of videos.Some other features: playlist import, translation, notes.I use the Transcript API to get transcripts for YouTube videos, and JustOneAPI for Bilibili videos. These services rely on the video's native transcripts. So this app does have this one limitation that if a video has no native

github.com• Apr 22, 2026• 1 min read

Show HN: Stonks-CLI – track your investment portfolio from your terminalHi HN,stonks-cli is a terminal investment portfolio tracker with live prices, P&L, stock financials, candlestick charts, news feed and AI chat botI built stonks-cli because I noticed I kept opening broker apps/TradingView/FinViz/Yahoo Finance multiple times a day just to check portfolio value, financial quotes, charts, different metrics, analyst insights and news.So I asked myself if I can have all this functionality in one app, preferably in in the terminal.What began as a small helper script turned into a terminal-first investment dashboard that now supports:* candlestick charts inside the terminal* stock fundamentals and analyst insights* crypto tracking* portfolio news feed* watchlists* importing portfolios from Interactive Brokers* an AI assistant that analyzes portfolio risk and news using RAGTry it with a demo portfolio (out of the box):$ pip install stonks-cli$ stonks demoI'd really appreciate feedback from people who manage portfolios from scripts, spreadsheets, or the termina

github.com• Apr 15, 2026• 1 min read

Show HN: Libretto – Making AI browser automations deterministicLibretto (https://libretto.sh) is a Skill+CLI that makes it easy for your coding agent to generate deterministic browser automations and debug existing ones. Key shift is going from “give an agent a prompt at runtime and hope it figures things out” to: “Use coding agents to generate real scripts you can inspect, run, and debug”.Here’s a demo: https://www.youtube.com/watch?v=0cDpIntmHAM. Docs start at https://libretto.sh/docs/get-started/introduction.We spent a year building and maintaining browser automations for EHR and payer portal integrations at our healthcare startup. Building these automations and debugging failed ones was incredibly time-consuming.There’s lots of tools that use runtime AI like Browseruse and Stagehand which we tried, but (1) they’re reliant on custom DOM parsing that's unreliable on older and complicated websites (including all of healthcare). Using a website’s internal network calls is faster and more reliable when possible. (2) They can be expensive since they

github.com• Apr 7, 2026• 1 min read

Show HN: Frontend-VisualQA — give coding agents eyes to verify their own UI workCoding agents today are blind.They write “valid” HTML/CSS code but can still ship a broken layout, a clipped dropdown, or a page at the wrong URL. Playwright scripts can assert modal.isVisible() without knowing the modal is rendered off-screen.Essentially, coding agents need “eyes” to verify their own UI work.frontend-visualqa is a CLI + MCP server for Claude Code and Codex for visual testing, verification, and QA of a website.You give it a URL and natural-language claims: frontend-visualqa verify http://localhost:8000/dashboard.html \ --claims \ 'The API status indicator shows Active' \ 'The monthly quota progress bar is completely filled' # → first claim passes, second fails (label says 100% but bar is ~65% full) It catches visualDOM disagreements that selectors are blind to.You can also test interactive flows without hardcoded data: frontend-visualqa verify 'http://localhost:8000/booking_form.html' \ --claims 'The date on the confirmation page matches the date selected on the calend

refearnapp.com• Apr 7, 2026• 1 min read

Show HN: RefearnApp – Open-source alternative to Rewardful/ToltI built RefearnApp as an open-source, self-hostable affiliate tracking tool you can run on your own infrastructure.It lets you track referrals and commissions, manage payouts, and keep full ownership of your financial and affiliate data. You can run everything on your own VPS and avoid third-party scripts in your app entirely.The goal is to give you full control over your affiliate system instead of relying on external platforms.I’d appreciate feedback on how you currently set up affiliate tracking in your products, and what problems you run into with existing tools.GitHub: https://github.com/ZAK123DSFDF/refearnapp

basecut.dev• Mar 31, 2026• 1 min read

Show HN: Basecut – share subsetted, sanitized prod snapshots with your dev teamMost teams get realistic data into staging one of two ways: write seed scripts that never quite capture production edge cases, or copy the real database and manually scrub emails and phone numbers by hand, hoping they don't miss anything (or, worse yet, not scrubbing at all). We built Basecut to replace both — automated snapshots with masking rules you define once, then restore sanitized copies anywhere. We're currently in early access and would love feedback from people who've dealt with this.

ai-proof-careers.com• Mar 27, 2026• 1 min read

Show HN: AI career guides for software career familyI created a set of career guides to help people adapt or pivot their career. The job market has been really tough for more entry level roles like "Computer Programmer" aka junior SWE or Web Developers. Job opening #s are dropping and employers are expecting more AI skills. So I tried to show things like which tasks are getting automated, what are hiring trends, and which emerging AI-native jobs ppl can try to pivot to.Full data and scripts are here, sorry they're super messy. https://github.com/yoyothesheep/ai-resilient-occupations-dat...What am I missing? Is this useful? Any feedback welcome!

news.ycombinator.com• Mar 25, 2026• 1 min read

Show HN: Herd – A Go sidecar to stop stateful processes Puppeteer/LLMs from OOMHey HN.I'm an engineering student at Waterloo building stateful AI agents, and I kept hitting the same wall: whenever my Python scripts crashed or dropped a connection, the underlying Puppeteer or Ollama processes would just sit there orphaned, eating RAM until the node OOM-killed itself. Standard load balancers break sticky sessions, and passive HTTP timeouts are too slow for cleanup.I couldn't find a good local process pool that actually cleaned up dead stateful sessions reliably, so I built Herd in Go.It uses a persistent stream (gRPC/Unix sockets) strictly as a dead-man's switch. If your client script dies, the stream breaks. Herd registers the EOF and instantly fires a SIGKILL to the worker process (relying on Pdeathsig on Linux). For the actual heavy data, you just blast HTTP traffic through Herd's internal proxy, which routes it directly to the active process port.My actual goal is to turn this into a multi-node distributed mesh with a Redis registry, where a client can drop off

news.ycombinator.com• Mar 22, 2026• 1 min read

Skills are quietly becoming the unit of agent knowledgeIn the last few months agent skills went from a niche Claude Code feature to something every major runtime supports. Anthropic has an official skills repo. OpenAI shipped skills in Codex with a built-in skill-creator. Karpathy talks about "everything is skill issue" and describes writing skills as curricula for agents [1]. The format is converging: a folder with a SKILL.md, optional scripts, optional reference files.What changed is that the models got good enough to follow written instructions reliably. A skill is just a tested workflow in markdown that the agent reads and follows instead of improvising. You can also bundle scripts that the agent runs during the workflow, which covers what most people use lightweight MCP servers for, except the agent can read the script source and extend it.Karpathy talks about an "economy of agents" and says we should stop writing HTML docs for humans and start writing markdown docs for agents [1]. Anthropic just shipped a skill-creator that benchmark

github.com• Mar 18, 2026• 1 min read

Show HN: PlanckClaw an AI agent in 6832 bytes of x86-64 assemblyI built an AI agent that fits in 6,832 bytes. The entire runtime (binary, bridges, tools, config) is ~23 KB.PlanckClaw is written in x86-64 assembly (obviously AI assisted code generation for this one) and uses only 7 Linux syscalls. No libc, no allocator, no runtime. The binary is a pure router: it reads messages from named pipes, asks another pipe what tools exist, builds a JSON prompt, writes it to a third pipe, parses the response, dispatches tool calls, and relays the answer. It never touches the network or executes tools directly.Everything else composes around it in shell scripts (~460 lines total): - bridge_brain.sh: curls the Anthropic API (~90 lines) - bridge_discord.sh: Discord Gateway via WebSocket (~180 lines) - bridge_cli.sh: terminal interface (~40 lines) - bridge_claw.sh: tool discovery and dispatch (~50 lines)Four processes, six named FIFOs, zero shared state. Adding a tool means dropping a shell script in claws/. No restart, no recompilation, no config change.It does