2025년 9월 2일의 Show HN

Check out my TUI I wrote for viewing and querying an LDAP. I need to do basic queries and validation daily for work, and as I work on a mac, there are really no good options. The major player is the Apache Directory Studio which is... not great. So I decided to create a new one.

Hi HN! Excited to share early access to Amber.

I’ve tried every all-in-one messenger out there (Beeper included) but they always fell short. No real folders, no AI, clunky UI, no CRM features… As a founder who speaks with hundreds of people every quarter, I needed something better. So I decided to rebuild the entire experience from ground up.

Thoughts from one of our users: "The app I've been searching for for a looong time."

Check it out for free today!

Demo: https://www.loom.com/share/8b5bc80b9893436b9190ae41fc3f0f50

Features: - All messages (Whatsapp, Telegram, iMessage) unified in a beautifully crafted interface. - Split inboxes (folders) to effortlessly focus on work, friends, a particular project.. whatever matters the most right now. - Mark read – no read receipts (even on Whatsapp and Telegram), mark done only when you’re done. - Personal CRM: a lightweight private database of knowledge about each person with AI pulling important facts straight from conversations. (the latter is coming soon) - Command bar + shortcuts. - Send later + reminders.

Everything is securely stored on-device. All messages are end-to-end encrypted and go straight from your device to the network of choice – never touching our servers.

Hi HN!

I'm Ruben, the founder of Davia (https://davia.ai/). Davia is a platform to build, edit and share applications, where builders get rewarded based on usage while users can discover and reuse any app's code to fit their own needs.

The problem we're trying to solve: Today's AI coding platforms charge you upfront for creation - it's like paying per block modification in no-code solutions. This works for commercial projects because you're aiming to generate revenue that exceeds your creation costs, but doesn't fit the reality of side projects and community apps that builders share for free. Most side projects require significant SEO effort and marketing to reach people who would actually find them useful, leaving creators with great apps but no discoverability. For users, when you need a specific tool for your project, you don't want to build it from scratch when someone has probably already created something similar.

Our solution is a YouTube-style platform for apps where builders publish with open code so anyone can view, modify, and reuse applications. Unlike traditional platforms like GitHub (designed for developers as builders and users) or app stores (where code is closed and can't be modified), we focus on making apps accessible and customizable for everyone.

We verify code quality and prevent spam to maintain platform standards. Builders get rewarded when their original apps are used as foundations for others' projects. Users discover apps, customize them for their needs, and build their own versions through reuse. They can either self-host or use our platform hosting, paying based on team usage when they scale to multiple users on our infrastructure.

You can try it here: https://davia.ai/. We'd love to hear from the HN community, whether you're a vibe-coder or curious about creator economics in the age of AI-assisted development!

I was experimenting with building a local dataset generator with deep research workflow a while back and that got me thinking. what if the same workflow could run on my own files instead of the internet. being able to query pdfs, docs or notes and get back a structured report sounded useful.

so I made a small terminal tool that does exactly that. I point it to local files like pdf, docx, txt or jpg. it extracts the text, splits it into chunks, runs semantic search, builds a structure from my query, and then writes out a markdown report section by section.

it feels like having a lightweight research assistant for my local file system. I have been trying it on papers, long reports and even scanned files and it already works better than I expected. repo - https://github.com/Datalore-ai/deepdoc

Currently citations are not implemented yet since this version was mainly to test the concept, I will be adding them soon and expand it further if you guys find it interesting.

StoryMotion, a hand-drawn motion graphics editor for educational content creators.

---

5 months ago, I launched Inscribed.app [1], a tool based on Excalidraw for making step-by-step explanation (like slides) that I can easily embed to explain concepts on my blog.

[1] https://news.ycombinator.com/item?id=43078555

as I later expanded into video, I decided to fork the project and create another tool which focus solely on making animation.

---

Key features:

- Keynote-liked interfaces with Excalidraw canvas - effects animation library - scene transition animation -- fun fact: you can also do “Magic Move” like you can do on Keynote :) - timeline editor for editing effects timing (with live preview) - use any fonts from Google Fonts

---

Try it yourself: https://storymotion.video/editor

I hope you like it. any feedback is appreciated.

Don't get your knickers in a knot. We were surprised too.

We are the team behind TideCloak which we shared on HN previously (https://news.ycombinator.com/item?id=42460131).

This time we ran an experiment. What happens when you apply a fundamentally different security model to the worst possible development process?

Behold... SecureAF https://secureaf.lovable.app/ Build video 6 mins https://youtu.be/tx5MsJ3jeQw

The code and UI are questionable and the quality assurance is non-existent. But the security model holds by treating even us as malicious, keeping authority over data out of reach for everyone except the rightful user.

Can you actually use SecureAF? Hell no. We built it in 20 minutes and never tested it. What's wrong with you?

What you reckon?

Hi HN, I’m Tobi. For a couple of hours over the past few days I’ve been hacking on something to fix a pain point in my dev workflow: writing status updates.

Progress is invisible by default. GitHub, Linear, Jira all track tickets and code, but they don’t do a good job of capturing the narrative between “ticket started” and “ticket done.”

You start working on a feature, your PM asks “how’s it going?”, and even though you know exactly how it’s going - because you’ve been committing and making progress - you still struggle to answer. That usually means breaking your flow to piece together an update, or just saying “it’s going fine.” You could point them to the commits, but tbj they probably don’t want to wade through diffs.

To solve this I built Zyg [pronounced zeig]. It tries to turn commits into human-readable progress updates. It’s a lightweight CLI + dashboard that wraps `git commit`. Running `zyg` will generate a detailed commit message from your changes, produce a project update from that commit or a set of commits you choose, and notify any stakeholders who are subscribed. If you’d rather not share updates automatically, you can just copy the generated summary and drop it in Slack or email.

Zyg is free for September thanks to an API credit grant from Anthropic. After that I’ll figure out pricing, but you can also plug in your own key and keep using it for free. It’s still rough around the edges, but I’d appreciate you giving it a spin.

I built a website featuring the paintings at https://pomological.art and a [https://chromewebstore.google.com/detail/mhoodlbpfiobhfbblii...](Chrome extension) that shows a random watercolor from the collection on each new tab.

*The story*: In the 1880s the USDA commissioned 7,500+ botanical illustrations to document American fruit varieties in a time when we had over 14,000 types of apples. The collection got digitized after decades in government storage (thanks especially to Kent Whealy and then Parker Higgins), but it remains scattered across government sites, Wikimedia, and the Internet Archive, and difficult to really explore and dig into.

*The site*: Lets you filter by crop type, browse by artist or range of years, examine particular varieties, or see where specimens came from on a map. You can explore how the same variety changed in appearance over time, or browse paintings that show plant diseases. My goal is to continue to research and start to provide history and context for the varieties themselves. The Chrome extension just serves up a random painting each time you open a tab. It's a nice way to stumble across varieties you've never heard of.

*Some context:* Previous HN discussions covered the collection itself (https://news.ycombinator.com/item?id=44287019, https://news.ycombinator.com/item?id=44397168). Parker Higgins FOIA'd the USDA and got them to drop their paywall after discovering they'd made a grand total of $565 over several years (https://parkerhiggins.net/2015/11/the-usda-pomological-water...).

*What's cool*: Apple hunters use these paintings to identify "extinct" varieties still growing in abandoned orchards. The Lost Apple Project matched fruit from a tree on Eastern Washington's Steptoe Butte to paintings of the Nero apple, which was assumed to be lost for a century. The Montezuma Orchard Restoration Project a few years back rediscovered the Colorado Orange the same way, matching fruit from a lone tree to four paintings in the collection.

*Why I care:* I'm starting a small cidery in Oregon (called Landrace Cider) and have become obsessed with rare and heirloom varieties over the years. I stumbled across the collection a couple years ago doing some variety research on [Orange Pippin](https://www.orangepippin.com/), and it still boggles my mind that such an incredible resource exists and lives in the public domain and yet so few people have ever heard of it.

*Also:*

I wrote a more in depth slash thoughtful essay on the topic, which tells the full story of the collection and gives more context for why the USDA felt compelled to pay artists to paint fruit: https://andrewhaupt.substack.com/p/a-field-guide-to-resurrec...

I built founder.mood after watching founders struggle with the pressure to always be "winning". The hard, lonely parts of building something never get discussed honestly because it can spook investors, customers, and employees.

So, I built a place with true anonymity by design, not just a hidden username.

The key technical choices:

No Social Graph: No profiles, no follows, no connection between posts.

Unlinkable Aliases: Every post gets a fresh, random name. Comment threads use a different names.

No Plain-Text Emails: Auth uses a non-reversible HMAC of your email. We can't email you, even if we wanted to.

Hard Delete: One-click account wipe. No backups, no retention.

The goal is to create a place for the kind of conversations that happen at 2:00 AM, not in a board meeting. It's post-first (no account needed) and built for safety, not engagement.

I launched this quietly and within the first hour, three founders posted about the loneliness of misunderstanding, the fear of failure, and the struggle of balancing work with life. It confirmed the need.

I'd love the HN community's thoughts on the approach—both the product and the architecture.

I built an MCP server that lets AI tools search Slack history to answer “why did we do X?” It reads messages and threads with your Slack user token, so agents like Claude Code/Desktop, VS Code and Cursor can pull context that isn’t in code or PRs.

Features: * Search by channel, user, date range, reactions, has:file, etc. * Get full thread replies. * Look up users by ID or display name.

Example uses: * “Why is foo_summary BigQuery table a materialized view?” * “Summarize this week’s topics in #random.”

More details in the blog post: https://dev.to/shibayu36/making-historical-context-easily-di...

We explain what Forward Error Correction (FEC) is, how and why it works in general, and how you can try it out with a new implementation in the Pion WebRTC stack.

This is a static type checker for mruby! It's still in development, but if you're interested, please take a look! We look forward to your feedback!

Hey HN,

In software development, the process of turning a user story into detailed documentation and actionable tasks is critical. However, this manual process can often be a source of inconsistency and a significant time investment. I was driven to see if I could streamline and elevate it.

I know this is a hot space, with big players like GitHub and Atlassian building integrated AI, and startups offering specialized platforms. My goal wasn't to compete with them, but to see what was possible by building a custom, "glass box" solution using the best tools for each part of the job, without being locked into a single ecosystem.

What makes this approach different is the flexibility and full control. Instead of a pre-packaged product, this is a resilient workflow built on Power Automate, which acts as the orchestrator for a sequence of API calls:

Five calls to the Gemini API for the core generation steps (requirements, tech spec, test strategy, etc.).

One call to an Azure OpenAI model to create vector embeddings of our codebase.

One call to Azure AI Search to perform the Retrieval-Augmented Generation (RAG). This was the key to getting context-aware, non-generic outputs. It reads our actual code to inform the technical spec and tasks.

A bunch of direct calls to the Azure DevOps REST API (using a PAT) to create the wiki pages and work items, since the standard connectors were a bit limited.

The biggest challenge was moving beyond simple prompts and engineering a resilient system. Forcing the final output into a rigid JSON schema instead of parsing text was a game-changer for reliability.

The result is a system that saves us hours on every story and produces remarkably consistent, high-quality documentation and tasks.

The full write-up with all the challenges, final prompts, and screenshots is in the linked blog post.

I’m here to answer any questions. Would love to hear your feedback and ideas!

This will work on any Hacker news post but it is especially useful for "Who's Hiring?". It will filter out comments that don't match the set filters.

You can set multiple filters. Filters have an "and" relationship. Filters are regular expressions.

Whodunit started as a pen-and-paper game for family game night. It evolved into a playable web app as the surprise reveal in a GopherCon '25 lightning talk.

The mysteries are all written by LLMs. Some are better than others, but most work nicely.

The web app is written in go with templ HTML templates, HTMX for on-page interactivity, and a turn-based game engine written as a Temporal workflow.

Check out the mystery gen write up for more info: https://blog.apartment304.com/whodunit-llm-murder-mysteries

We kept breaking production with small prompt edits — suddenly outputs weren’t valid JSON, fields disappeared, or formats changed silently. So we built Promptproof, a GitHub Action that runs in CI and blocks PRs when prompts produce invalid outputs.

Features:

- Validates JSON output

- Enforces required keys & schemas

- Runs fast in CI (no external infra)

- Works with OpenAI, Anthropic, and local models

- Adds PR comments so reviewers see failures immediately

We’d love feedback: which rules or integrations would make this most useful for you?

Repo: https://github.com/geminimir/promptproof-action

Hi HN, I’m Tina Over the past months I’ve been frustrated with how often LLMs hallucinate — they generate answers with high confidence, but sometimes the information simply doesn’t exist. For teams relying on AI (content, research, finance, even legal), this can be a serious problem. So I’ve been working on CompareGPT, a tool designed to make LLM outputs more trustworthy. Key features: Confidence scoring → surfaces how reliable an answer is. Source validation → checks whether data can be backed by references. Side-by-side comparison → run the same query across multiple models and see consistency. You can try it here: https://test.comparegpt.io/home Right now it works best in knowledge-based queries (finance, law, science), and we’re still improving edge cases. I’d love to hear your thoughts, feedback, or even brutal criticism from this community. Thanks for reading!

I created a public API collection repository because I was tired of searching for open APIs all over the place while working on projects.

It includes the latest (2025) English translations of Korean open APIs.

To keep the links fresh, I also added a routine that regularly checks them, so everything stays stable and reliable.

Feel free to take a look around, and any feedback is always welcome.

Thanks!

Hey HN I want to expose what is happening with Stripe specialist team scam and I am telling it as someone who had $800k+ held and got offered 72% deal to get it released from someone inside the Stripe and actually has seen them to release my friend's account with my own eyes after 2 days of accepting the deal

In short, they make you beg for your own money and never release it and then in groups in telegram and other places offer 50%+ deals (my own deal was 72% for $800k held) where they take huge cut from released funds.

Proof: I have my friend who was desperately trying to recover his funds while having absolutely no dispute and 2 days after taking the deal, his funds automatically get released

So what is happening is SOMEONE or SOME GROUP in specialist team running a SCAM where they are holding someone's funds forever and basically create enough chaos where people are desperate.

Then they have a group advertising a way to release funds for 50%+ deals, if you do not take it funds never get released basically 2-3 days after you accept the deal, you give them your account ownership, they send funds to their own bank and pay you your share (exactly what happened with my friend)

How exactly do they make you desperate: 1) You contact chat 2) They tell you bear with me while I am looking into your account 3) They explicitly make you wait around 40-50 mins so you don't try again 4) They say they will need to convert this chat to email because only specialist team can handle these cases 5) Once chat converted into email, they completely ignore you 6) In the dashboard, it is shown TICKET IS CLOSED few mins after getting opening

I have done this for last 9 months as I REFUSE to give up 72% of my funds for criminal scams. I have 0 disputes. We were using Stripe Standard Connect to help companies to get customers through email marketing, and ads and charge % of sales and we were specifically using direct charges which means we are not liable for disputes of connected accounts either

--- When you report this to the chat saying do not convert it to email, they assure you this time they will send it to CORRECT team or this team they will send it to compliance team, or higher ups but next business min after you create the ticket it will get closed

Oh also, we sent legal letters and they are still ignoring it. I am now trying to find lawyers to help me to sue Stripe and release funds for the % of my funds (probably a lot lower than what Stripe Specialist Scam team offered us)

I am trying to expose this scam because MAYBE for the sake of their reputation they will do something and I would be happy to be quite because I only care about my funds being released. My entire business got shut down after closing my account

Hi HN,

I think mesh gradients are cool, I use them a lot, probably too much.

So I made a simple tool that puts one behind my screenshots.

No watermark, no sign-up, I just paste my screenshot, select my current favorite color palette, I add an unhealthy amount of grain texture, and hit "copy". Lots of aspect ratios available.

I also made a Canvas tool to make more advanced designs with optionally transparent background.

Do check it out and tell me what you think