PH热榜 | 2026-02-27

Hey all,

I'm Kiet, one of the creators of Superset. We created Superset to reduce the time you spend waiting around for agents like Claude Code to run. Superset lets you kick off dozens of coding agent sessions in parallel. It works with any coding agents like Claude Code, Codex, OpenCode, etc. while allowing you to use your Pro/Max plans. It's a full featured IDE optimized for parallel agents workflows.

The last few months have seen incredible adoption from the most cutting edge teams from all over the world. I'm excited to see what you will build with Superset!

Hey there,

I'm Satya, I'm also part of the core team building Superset! It's been amazing seeing our product grow over the last few months up to this 1.0 launch. We've helped so many people already improve their Claude Code, Codex, etc. CLI flows by helping them manage worktrees in parallel, organize Linear tickets and more. It's been really gratifying to watch the product grow, I just want to shout out all of our users - they've been incredibly supportive and helped us get to a product we're all very proud of.

If you end up using Superset from this post I'd love to hear what you think!

The sandbox isolation is the right architectural call to prevent interference, but it creates the inverse challenge: when two agents are working on related tasks — say, a frontend and a backend agent that both need to agree on an API shape — how does Superset handle cross-agent context sharing? Full isolation is clean but can lead to agents making conflicting assumptions in silence. Is there a shared workspace layer agents can read from, or is isolation by design absolute?

Hey yall,
I'm Avi, one of the creators of Superset. I'm super excited to share what we've been building. We use Superset to build Superset and its really fun to see how much people are enjoying it. Theres a lot more we are planning on shipping but I'm looking forward to hearing all the feedback from more users, and to see what you all think of it!
Let me know if you have any questions :)

love this, it’s my daily driver now. replaced cursor, claude code, and conductor!

Let's gooo massive fan of Superset product + team! One of the fastest executing teams I've seen!

Question for you guys: Do you think the future will see more local or cloud development? Right now I'm using my mac mini and ssh'ing into it, and it's great in many ways but can get messy with enough things running in parallel. Cloud is easier to manage but problem becomes continuing working from the CLI and all that. Where do you think things are heading?

This looks really solid love the idea of running multiple agents in parallel

Quick question — with all these isolated environments, how are you handling security between them?

Like making sure one agent can’t access another or any API-related risks?

Been a power user. These folks ship so fast I get an update every day!

Congrats on the launch. This is super cool! Are agents aware of how much ram they are using on the host machine to prevent slowness?

Interesting approach to running multiple agents locally. Curious, how are you thinking about agent coordination to avoid conflicting code changes?

Really love that this works with existing Pro/Max plans, that’s a huge unlock for teams already deep into AI coding workflows. Such a smart move 🙌

Big fan of this team and product. We use it daily. Congrats on the launch!

Running multiple coding agents in parallel without context switching is exactly where dev workflows are heading. The sandbox isolation + centralized monitoring feels like a serious productivity unlock, especially for teams experimenting with Claude, Codex, and other agents simultaneously.

The built-in diff + review layer is a smart touch too. That’s usually where agent workflows get messy.

Curious:

• How does Superset handle long-running agent memory?

• Any benchmarks on productivity gains vs. single-agent workflows?

• Is there team collaboration support on the roadmap?

This solves a real pain point. I've been running Claude Code sessions sequentially and the context switching kills momentum. The sandbox isolation per task is smart — I've definitely had agents step on each other's changes when working on related files.

Question: how does the diff viewer handle conflicts when two agents modify overlapping files? Is there a merge flow or does it flag it for manual resolution?

Great launch, congrats to the team!

Great ease working with claude code etc using SuperSet!! No more headache and waste of time!

The waiting problem with Claude Code is real, but I'm curious how Superset handles the orchestration layer. When you have dozens of agents running in parallel, how do you decide which tasks are safe to run concurrently versus which ones need to be sequenced to avoid conflicts?

And with sandbox isolation per task, does each workspace get its own file system snapshot? Wondering how state gets reconciled if two agents end up touching related parts of the same codebase.

I use Claude Code daily for building my AI platform and parallel agent workflow conflicts are something I've been thinking hard about. Would love to understand more about how Superset handles those edge cases in practice. Congrats on the launch!

Congrats on reaching #1 today, Superset team! The ability to run multiple coding agents without context switching is a game-changer for speed. As a developer building a PropTech SaaS (ParkEase) with React Native and Python, I can see how this would have saved us tons of hours during our latest sprint. Question: Does it support custom sandboxes for private Python libraries? Good luck with the launch!

Finally an IDE for people who find waiting 30 seconds for Claude Code physically painful. Now I can be unproductive in 12 parallel threads instead of one 🙏

already use yall daily haha

Worktree isolation per agent is the key design choice here. Without it, parallel agents would constantly step on each other's changes. The persistent daemon surviving crashes is a nice touch too. How are you handling merge conflicts when multiple agents edit overlapping files?

Another IDE, of which there are already many. How are you better than the others?

Been using superset since the beginning and it's great for managing multiple projects!

Congrats on the launch! As someone who often uses just a single aider session at a time and calls it a day, very excited to try this out!

Been using superset consistently for the last month and my usage on it is probably > cursor right now! Insanely helpful tool if you're trying to work parallel coding sessions (multi projects, worktrees, etc.)

Does this handle git worktrees automatically? Do I need to reclone all my existing repos as bare repos in order to use this?

Congrats on the launch!

Anthropic announced a new Claude Code feature called Remote Control.

The idea is pretty straightforward: you start a Claude Code session locally in your terminal, then continue them from your phone. Take a walk, see the sun, walk your dog without losing your flow. It's rolling out now to Max users as a research preview.

Anyone here on Max plan tried it yet? How the mobile experience feels in practice (latency, editing capabilities, etc.)?

"Take a walk, see the sun, walk your dog without losing your flow" is doing a lot of work in that description and it genuinely lands. The mental cost of losing context mid-session is way higher than people realize until they experience picking up exactly where they left off.

The one remote session at a time limitation is an interesting design choice. Keeps things simple and avoids conflict, but I'm curious if that becomes a friction point when you have multiple long-running jobs going and want to check in on a specific one from your phone.

I use Claude Code daily for building my AI platform and the number of times I've walked away from my desk mid-session wishing I could nudge it or check progress from my phone is higher than I'd like to admit. This is exactly the feature I didn't know was missing.

This is exactly what I always had in my bucket of TODOs, while I travel in bus to University and the way back is in total 3 hours and I believed having something like this would seriously make that time of mine productive.

Been using the @Cursor Agents, @Jules and recently the @OpenClaw to work in that time of the day. This will be the next try but I always thought how about having a computer on the go with me in the bus (not literally but to be able to control one remotely).

Visualising this I had one image like being able to project through maybe some VR glasses and do it? But to be cheap I also thought about how using a mobile maybe we can remotely do this and we're here.

What you'll think about the future ? Let me know and I'll share something I truly believe is the future but so much time for it to happen it seems.

Congrats on the launch! Being able to monitor local sessions from a phone is a huge win for mobility. As we develop ParkEase (a SaaS for property managers), we’re constantly looking for ways to make complex backend tasks manageable from a mobile device.

Quick question: Does the remote control interface support multi-tenant sessions, or is it strictly one agent per remote link? This looks super useful for our dev workflow. Good luck today!

Nice for monitoring long-running tasks. Would be good to get push notifications when it needs approval or hits an error. Or you already support that?

Remote access to local sessions is the feature I've been waiting for — I do a lot of quick steering of Claude Code runs while away from my desk, and the context loss when stepping back is the main friction. Curious about the interaction model: is it full bidirectional control (add instructions, see live output, approve/reject steps), or more read-only monitoring with an interrupt? That distinction changes a lot about how you'd actually use it in a 'ship while mobile' workflow.

Super excited about this launch! :)

Perplexity Computer feels like a real step beyond “AI that chats” into AI that actually works.

Here’s why it’s interesting: It doesn’t just suggest. It executes.

• Breaks your goals into tasks and subtasks

• Spins up specialized sub-agents

• Runs them in parallel

• Keeps working for hours (or even months) until the job is done

It can handle:

• Web research

• Document drafting

• Data processing

• API calls to your connected tools

All inside isolated compute environments with:

• A real filesystem

• A real browser

• Real tool integrations

That’s a big leap from typical “assistant” products.

Massively multi-model by design

Instead of relying on a single LLM, it:

• Uses a core reasoning engine

• Delegates subtasks to specialized models

• Routes each step to the best model for the job

Research → one model
Images/video → another
Speed tasks → another
Long-context recall → another

It’s model-agnostic and gives users control over model selection + token budgets. As models evolve, your system evolves with them.

The name actually makes sense

Historically, a “computer” was a person who divided and completed complex calculations.

This feels like the modern version: A system that coordinates powerful models across tasks, tools, and time with accuracy as a core requirement.

Built on the same foundations as Comet (AI-native browser) and Comet Assistant. Available for Max subscribers.

Feels like Perplexity moving from “answers” to full, end-to-end execution.

The "hours or even months" claim is where it gets genuinely hard — most multi-agent systems don't fail on capability, they fail at decision points where the right move is ambiguous and the system either halts or guesses wrong at a cost. Curious what the governance model looks like for those moments: does Computer surface a checkpoint when it hits an underspecified branch mid-task, or does it attempt to resolve autonomously and flag it in a post-task summary?

This feels like a big step toward AI systems acting more like full project operators instead of just assistants. The idea of orchestrating multiple models automatically is really interesting.

Moving from answers to end-to-end execution is a bold step. Huge congrats on the launch, this is exciting to watch 🔥

The historical framing of "computer" as a person who divided and completed complex calculations is spot on and it reframes what this category of product actually is. Not an assistant. A coordinator.

The massively multi-model architecture is what separates this from everything else launching in this space. Routing each subtask to the best specialized model rather than forcing one LLM to do everything is the right call, and it mirrors how serious AI pipelines are actually built today.

I build multi-model orchestration systems myself and the hardest problem is always the routing layer, knowing which model to trust for which task and when to hand off. Curious how Perplexity handles that decision in practice, rule-based or learned? Congrats on a genuinely ambitious launch!

Congrats on the launch! What you have done by integrating every current AI capability into one system is Awesome. This makes Perplexity Computer that more efficient.

Quick question: Regarding errors in coding, how does Perplexity Computer tackle errors, does it have a specific AI Agent in which its sole function is to seek errors and correct them?

I also would love to get your feedback on my platform just built, from your perspective of everything that Perplexity Computer does and how it functions. Would love your feedback if you get a chance!

Oh wow, there's really no demo/trial and it pushes you to paying 2K? :-O

whenever i see perplexity i feel prouf of you @aravindsrinivas

Major update from @Google: Nano Banana 2 🍌 (Gemini 3.1 Flash Image). It’s state-of-the-art for creating and editing images, combining Pro-level capabilities with lightning-fast speed.

You now get:

• Advanced world knowledge powered by Gemini

• Real-time web search grounding for accurate subject rendering

• Rapid edits and iteration with Gemini Flash speed

This isn’t just faster images, it’s context-aware generation.

What's new?

1. Stronger instruction following: Better adherence to complex prompts means fewer retries and more predictable outputs critical for production workflows.

2. Accurate text + localization: Clean, legible in-image text and translation & localization support. Huge for marketing creatives, infographics, greeting cards, and global campaigns.

3. Subject consistency: Maintain resemblance across up to 5 characters. Preserve fidelity of up to 14 objects. That unlocks storyboarding, ad sequences, brand mascots, and UI continuity.

4. Production-ready visuals: Richer lighting, textures, sharper details. Flexible aspect ratios. Resolution control from 512px to 4K. No more heavy post-processing or upscaling hacks.

Rolling out across:

• Google Ads

• Google Antigravity

• Vertex AI (preview)

• Search AI Mode & Lens

• Flow (default, zero credits)

• AI Studio + Gemini API (preview)

• Gemini app (replacing Pro in Fast/Thinking/Pro modes; Pro still accessible via regenerate for AI Pro & Ultra users)

Expanded to 141 countries + 8 new languages. This isn’t a demo model, it’s deeply embedded.

Provenance built in

• SynthID watermarking

• C2PA Content Credentials support coming to Gemini

• 20M+ SynthID verifications already

Google is clearly building long-term trust infrastructure.

It's blazing fast. Masters text in any language. Handles custom aspect ratios, generates up to 6 consistent character images, leverages real-world knowledge for hyper-niche visuals, and taps real-time web search data for spot-on renders of anything specific.

This is clearly aimed at high-frequency pipelines, think bulk UGC ad creation, or generating consistent frames for video models like @KLING AI or @Seedance AI Video Generator. If the pricing is as low as the previous Flash models, this might be the most important release.

What do you think? :)

Subject consistency across up to 5 characters and 14 objects is the feature that jumps out for me. That's the capability that makes brand mascots, ad sequences and storyboarding actually viable at production scale, without manual frame-by-frame fixing.

The accurate in-image text with localization support is also a big deal for marketing creatives. That's been a persistent weak spot across every image model and getting it right unlocks a huge amount of real campaign work.

I orchestrate multiple image models in my AI branding platform and Flash-speed generation with production-ready quality is exactly the kind of primitive that changes what's economically feasible in high-volume creative workflows. Curious how it holds up on complex typographic layouts. Congrats on the launch! 🍌

Google is moving crazy fast here.

Honestly, competitors still haven’t caught up with Nano Banana Pro, and now they shipped another update already.

The iteration speed is kinda wild — feels like they’re widening the gap every release.

How does Nano Banana 2 perform compared to other leading image models in terms of realism and prompt accuracy?

Never compacts’ isn’t a feature, it’s a mentality. Most agents lose the plot like a junior pushing to prod on Friday at 5pm. This isn’t about cute summaries, it’s about decision-grade memory. If you truly preserve those rare constraints from three hours ago, that’s a real shift. Everything else is just context theater.

wow this is just AWESOME!

We use MastraCode internally to build Superset and it’s 🔥

The team ships lightning quick and has great taste!

"Never compacts" is the right framing — the real failure mode is mid-session context amnesia, not the obvious stuff. The hard part is that standard summarizers preserve what's recent or frequent, but the detail that derails a session is usually the rare constraint mentioned once: a deprecated API, a specific architectural decision made three hours ago. Does the observational memory layer use explicit signals (test failures, user corrections) to weight those rare-but-critical details, or is importance inference more heuristic?

Paul from Mastra here, excited to launch Mastra Code 🎉

Mastra Code is a CLI-based coding agent built on the observational memory feature we shipped earlier this month. Our team uses it as their daily coding driver, and we think you’ll love it.

What problem does Mastra Code solve?

A lot of coding agents feel great at first, until the context window fills up. When this happens, the agent uses compaction to summarize past conversations, often dropping important details and forcing you to repeat work, re-explain context, or double-check what its remembered, which can become really frustrating.

Mastra Code is different. It's powered by our state-of-the-art observational memory, which watches your conversations, generates observations, and reflects on them to compress context without losing important information.

The result is simple: long-running coding sessions that remember what matters so you can build faster, merge sooner, and ship more.

Get started

Install mastracode globally with your package manager of choice:

npm install -g mastracode

If you prefer not to install packages globally, you can use npx:

npx mastracode

What it's like to use

Mastra Code runs directly in your terminal, with no browser or IDE plugin required. With most coding agents, you spend time managing context windows, splitting work across threads, or saving notes before compaction hits. With Mastra Code, there is no compaction pause and no noticeable degradation, even in long-running sessions.

No compaction! Even with 1M context window compaction took like 3 minutes. With Mastra Code I don't notice any degradation, I don't curse into the air, I stopped yelling COMPACTION, and my mental health is better for it. — Abhi Aiyer

You can throw anything at it: planning, brainstorming, web or code research, writing features, and fixing bugs. Over time, memory fades into the background so you can focus on building.

I don't worry about the conversation length or multiple threads for anything. I just keep rolling and it keeps going. — Daniel Lew

After a few days, you realize you're no longer tracking context windows or restructuring work to avoid compaction. Observational memory quietly remembers what matters as sessions grow. Once you experience that, it is hard to go back.

We'd love your feedback. Drop questions below and we will be here answering all day.

This is really interesting the “no compaction” part hits hard, that’s a real pain

Curious — with long-running sessions and persistent memory, how are you thinking about security around stored context?

Like preventing sensitive data leaks or unintended access over time?

I have been using MastraCode daily for many weeks now!

Things I love about using it:

1. Never hitting the break in momentum from compaction

2. Long single sessions were an anti pattern but now I do them all the time and my agent recalls nicely

3. You can build your own version of it using Mastra OSS Harness!

The observational memory approach is really compelling. Context compaction has been my biggest frustration with long coding sessions — you lose that one architectural decision from 2 hours ago and suddenly the agent is working against your own codebase.

Question: how does the memory layer handle conflicting information? E.g., if early in a session you say "use REST" but later switch to "actually, let's go with GraphQL" — does it pick up on the correction or does the older observation persist?

Congrats on the launch!

@marshallk This is really cool! Can you ask follow up questions as well if you need more insight or clarification? Are there one-click save options like export as PDF or an excel file?

This is dangerous in a good way. Most people skim and feel smart. This actually forces structured thinking on top of what you read. The ‘state of the art map’ angle is strong if it truly separates signal from noise in 10 seconds, that’s serious leverage.

Curious how you prevent analysis overload with 35+ tools though. Power is great but clarity wins.

I've been able to test drive this prior to launch, and it's easily become part of my workflow. Quick and solid context for whatever I'm digging into. Well done, @marshallk !

I am using the extension on Edge, but it didn't work. I thought the extension would automatically detect the page and analyze it, however, it asked for a local PDF, and even after uploading one, it didn't return any results. Maybe it's a browser compatibility issue?

The mental models angle is a really clever differentiator, Marshall. Most AI reading tools just summarize — this actually gives you frameworks to think about what you're reading, which is way more useful for decision-making. As someone who builds browser extensions too, I'm curious how you handle the context window when analyzing longer articles. Do you chunk the content or send the full page to the model? Also, the "35 AI tools" part is interesting — are these pre-built prompts layered on top of the article content, or something more dynamic? Would love to understand the architecture behind that.

The HN website is known to be "minimal", to say the least. Why might that be the case for so long, when they clearly have the ability to make it look more modern? Maybe for more classic look? idk but I am very relieved to have a more modern HN.

Is this just me??

Love that you open-sourced this, Dylan. The visual grid with Open Graph thumbnails is such an obvious improvement over the default HN wall of text — surprised nobody built this sooner. The side-by-side article + comments view is the feature I'd use most. Half the value of HN is the discussion, but switching between the article and the comment thread in a browser is clunky. Having them next to each other is the right UX. Quick question — what made you choose SwiftUI over something cross-platform? Was it purely about getting that native macOS feel, or were there specific SwiftUI features that made the visual grid easier to build?

This is a really cool idea turning daily data into a podcast is super unique

Curious — since you're pulling data from multiple sources (calendar, GitHub, etc.), how are you handling security & access between these integrations?

Especially making sure everything stays isolated and safe

Hey @marcfinger this is a really compelling concept. Three questions on ownership and portability: Do users own the underlying audio files outright? Can episodes be downloaded and repurposed across other channels? Is the RSS feed configurable to be public vs private?

From a founder perspective, this could be a powerful always-on content engine if the outputs are fully portable.

Interesting concept! What other integrations are you planning to add soon?

Lemonpod making a podcast about your life where the host says "and on Tuesday you committed 0 lines of code but took 3 naps" with full journalistic gravitas 💀

Congrats for the launch!

Love the idea of turning daily data into a podcast. Have you seen users stick with it long-term, or is it more of a novelty at first?

Love the concept :) as a lover of podcasts I'm always looking for sth new... Does the RSS feed work with any standard podcast player (Spotify??) or is it optimized for the app?

Cool idea, personal morning podcast is smart. Сan I control how long it is and how detailed it gets? Would love Gmail or Notion integration too.

Hey Product Hunt 👋

I’m Connor, founder of Alkemi.

We started Alkemi to remove the gap between data and decisions.

We built the Alkemi Slack Agent around a simple observation:

Decisions happen in Slack.

Data does not.

Every company claims to be data-driven. The data exists, but little of it informs their actual day-to-day decisions because extracting answers is too slow and expensive. Essentially, the juice isn't worth the squeeze in most cases. BI teams, dashboards, and monolithic tools only add layers of complexity.

So people make suboptimal decisions.

The Alkemi Slack Agent turns Slack into your team’s on-demand data analyst.

Ask questions in plain English.

Get charts, summaries, and answers instantly, powered by Snowflake, HubSpot, BigQuery, Databricks, and more.

No dashboards.

No backlog.

No more asking for a quick favor.

Just:

“How should I reallocate marketing spend?”

“What drove pipeline last week?”

“What products have untapped revenue potential?"

And you get a real answer, grounded in your data.

Under the hood, we convert data sources into structured, agent-usable functions. That’s why the responses are reliable, not just convenient.

Our belief is simple:

If everyone had a data scientist in their pocket, with zero waiting and zero political cost, decision quality across the company would materially improve.

This launch is about one thing:

Collapsing that space, and bringing data directly to the moment of decision.

We want blunt feedback.

Let us know what would make this essential for you.

We’re shipping fast.

Thanks for taking a look.

I live in Slack, but I still spend so much time navigating external systems to get the answers I need. That kind of ambling around BI tools, native ad tools, spreadsheets, etc... is such a time suck.

Early Alkemi fangirl here - love how you're stitching proprietary data together in the answer-engine format that is becoming my everyday search format.

Congrats, super practical. How do you make sure people only see the data they’re allowed to see? And what happens if the AI gives a wrong or incomplete answer?

Meet MaxClaw: One-click deploy of always-on agents in Telegram/Discord/Slack/etc. Powered by M2.5, persistent memory, personalization, zero maintenance/servers.

Love the helix DB team and product. Congrats!

Graph + vector in a single engine built in Rust is exactly what the agent ecosystem needs right now. Most people are duct-taping Postgres + Pinecone together — having native support for both traversal and similarity search in one place should make agentic workflows way cleaner. Excited to see what the HelixQL language evolves into.

Really impressive journey especially scaling to billions of queries

Curious — with this kind of workload, how are you handling security around data access and isolation?

Especially if teams are using it in multi-tenant or production environments

Graph + vector + OLTP in one engine is interesting. Are you targeting agent memory use cases primarily, or positioning this as a general-purpose database long term?

Hi Product Hunt! I'm Joon from MVNT.

We believe everyone deserves dance - so we are building the world's best dance gen model! While the tech is a long-term journey, we just couldn't wait for user validation and feedback 😛

So we built a simple, fast, funny web-based playground where anyone can play around with dance with ease! Here's what's inside v0.1:

- Music to Dance: paste a YouTube link or upload a file to get a dance in under 3 minutes. (Powered by our diffusion-based model mvnt-m4)

- Image to Dance: upload any character image and bring them to life dancing. Powered by Tripo AI's v3 model.

- Screen Recording: export your dance as video. Dance along to it, or use it as motion reference for tools like Kling AI or Seedance.

- Community Gallery: the funniest part. See what other creators are generating in real time.

We've long been a research-focused team just getting started with product. We are still working on better dance quality, faster generation, MP4 downloads, finger & facial motion.

Come join our playground, and thanks for being with us from day one!

Discord: https://discord.gg/hpWFbjb6

LinkedIn: https://www.linkedin.com/in/jooooooon-jung/

running a small team myself, and really looking forward to trying this out!

This looks really powerful, I can see the AI providing a lot of value for conversations around root cause analysis and planning

한국어로 패치할 계획이 있나요?

Why this over something like Claude’s workspace

This hits a real nerve. I've lost count of how many bug reports I've gotten that just say "it doesn't work" with a screenshot of a blank page. Having non-technical team members capture actual request context without opening DevTools is a smart approach.

Question: do you plan to support auto-redaction of auth tokens/sensitive headers before export? That would make it much easier for support teams to share captures without worrying about leaking credentials.

Nice launch, congrats!

Congrats on the launch, Victor! As a fellow Chrome extension builder, I really appreciate how you've identified such a specific pain point. The "it's not working" bug report with zero context is something every dev has dealt with, and the idea of bridging that gap for non-technical team members is smart positioning. Making network request capture accessible without DevTools is a great angle — it removes the friction that stops QA and support teams from providing useful data. Curious about your approach to the Chrome extension review process — did you run into any challenges with permissions for intercepting network requests? That's always been a tricky area for extensions in this space.

Quick update — I’ve just added an English version of the demo video to make it easier for everyone to follow along.

The per-tenant ML model approach is really interesting, Ilyas. Most security tools apply the same generic rules to everyone, which is exactly why alert fatigue is such a problem — everything looks like a threat when you don't know what's normal for a specific environment. The fact that each customer gets their own baseline model that evolves over time is a strong technical moat. Starting with Cloudflare Enterprise is smart too — you're going where the security-conscious customers already are. How long does the model typically need to establish a reliable baseline for a new customer? That ramp-up period seems like it could be a key factor in onboarding experience.

Congrats with the launch! Do you see this becoming more of a curated ‘shopping assistant’ or more of a performance layering guide (or both)?

I'm just about to go out for my run in -2°C, this came in handy. Cool idea.

Congrats on the launch!

UPDATE: Feb 27 2026

I added MiDI interface for MusiKey so individuals can hook their instruments to it.

MusiKey — Musical Entropy Authentication

MusiKey replaces passwords with music. Instead of memorizing strings of characters, you authenticate with a unique musical composition that doubles as a cryptographic key.

The problem: Passwords are weak, reused, and forgettable. Hardware tokens are expensive and losable. Biometrics can't be changed if compromised.

The solution: MusiKey generates a random musical composition tied to your passphrase. The composition itself becomes your cryptographic credential — something you can hear, recognize, and verify, but that carries 112+ bits of entropy. If compromised, just generate a new song.

What it does:

• Generates unique musical compositions as authentication credentials

• Encrypts them with cascaded KDF (PBKDF2 + Argon2id) and double AES-256-GCM

• Acts as an authenticator for external services via ECDSA P-256 signed challenge-response (MusiKey Protocol) — a more secure alternative to TOTP codes

• Supports multi-factor auth: musical challenge-response + time-based codes

• Visual fingerprint lets you visually confirm your credential at a glance

• Self-destructs after 5 failed attempts — no brute forcing

• Full audit log with tamper detection

What it isn't: MusiKey isn't a password manager or a music app. It's a proof of concept that musical structure can carry enough entropy to be cryptographically useful, while being more human-recognizable than a random string.

All processing is local. Nothing leaves your machine. Zero runtime dependencies. Open source.

Update on Feb 27 , 2026: MusiKey turns music into cryptographic keys. Generate a unique composition, and it becomes your authentication credential — encrypted with cascaded KDF (PBKDF2 + Argon2id) and double AES-256-GCM. Features ECDSA P-256 challenge-response authentication for external services, multi-factor auth (musical challenge-response + TOTP), visual fingerprint verification, tamper-detected audit logging, and self-destructing credentials after failed attempts. Zero runtime dependencies — all crypto via platform APIs. Open source, fully local, nothing leaves your machine.

Fascinating idea. It could do with a little more explanation and UI tweaks - I could generate compositions but couldn't work out if there were other things it should also be able to do.

Also I think the shown keyboard is an octave shorter than the range of the generated notes (low notes appeared on the indicator about the keyboard, but not on the keyboard itself).

Edit to add: Also this could be deployed directly as a web app for demo purposes (using browser storage instead of file storage), rather than getting people to download and install a git repo.

Congrats on the launch! Shipping a desktop app with solid auth is no small feat. I saw the GitHub repo, what license are you plan to use (MIT, etc.)?

Love the 'chip-on-the-shoulder' energy behind this—solving a personal frustration with a data-driven solution. Verifying through GitHub and Stripe history is a much more modern approach to trust than just a passport.

What I love about this is the founder story - 19 years old, watching his father get rejected, builds the infrastructure to fix it. That's the kind of chip-on-your-shoulder energy that actually changes industries. Following closely.

This hits close to home. I've watched talented devs from restricted regions build real products, real revenue, real track records — and still get denied by a bank that never looked past their passport. The idea of using your actual digital work history as proof of legitimacy is long overdue. Built by founders who lived it shows — this is the kind of problem only insiders solve properly.

Hey Product Hunt community 👋

I’m Pavel, the founder of Honoramma.

After 10 years, Honoramma is finally launching on Product Hunt. Honestly, the concept scared me at first, but I kept coming back to it. I once heard Sam Altman say that the best ideas often sound “terrible” at first. Maybe this is one of them 🙂.

The problem

Even in 2026, we still don’t have a modern digital space for memory. The main place we go to remember someone is still an offline memorial.

This may sound harsh, but since Facebook launched and social platforms took over our lives, more than 1 billion people have passed away. And what have we, as the digital world, built for memory? What will happen to our social profiles?

Yes, we have databases of the deceased and family tree tools, but those are information organizers, not ways to express memory and relationship. Feeds are great for news and updates, but the atmosphere of remembrance, I believe, needs a different format.

Our solution

Honoramma is a platform for building an entire world of memory and respect. It’s a place where anyone can create a monument for a person, an event, or a pet, or build their own park and connect relatives’ monuments into a family tree.

We go beyond the traditional memorial format by focusing on the expression of honor, whether someone is living or not. These are places of memory, and they exist outside of time, especially in our fast-moving world where almost everything flows away like a river.

What is it for?

✨ Gather your family history into one shared park

✨ Express respect for public figures

✨ Honor meaningful eras and events

✨ Create a single monument for someone important

I made a few example parks:

👉 Bach Family Park 👈 complete family tree
👉 Beyond Earth 👈 about space
👉 Leonardo da Vinci 👈 a single memorial

Something unexpected happened

One of our early users created a park called👉Launch Stars Park👈a tribute to iconic product launches that inspired them. And somewhere in that park, there’s a very small Honoramma monument standing next to much bigger ones.

That made me smile 🙂Huge thanks to El for creating it.

How does pricing work?

Even though subscriptions are popular today, I don’t think that model fits “eternal memory.” What happens if there’s no one left to keep paying? That’s why buying a monument slot or a park map is a one-time purchase.

You don’t need to pay anything to explore or try Honoramma. Payment is only required when you decide to activate your own park.

Here are the starting prices:

✨ Small park: $55
✨ Medium: $95
✨ Large: $215

🚀 First-week launch credit available: we cover $200 of your first 30×30 Park. You pay $15.

Monument slot prices are set by the park creator, but the current default prices are:

✨ Small slot 1×1: $19
✨ Medium slot 2×1: $29
✨ Large slot 2×2: $99

For example, for our Product Hunt launch, I created a special free park 👉 Free Launch Park 👈 where monument slots are free.

Because there’s no subscription model, sustainability becomes the key question.

We built Honoramma around paid tributes (small paid actions inside parks), and we believe that ongoing flow will be enough to keep the platform healthy and running.

Curious what you think 😉 I’ll be here all day, ask me anything

If you're curious about long-term sustainability, data permanence, or how ownership works, I’ve added detailed answers in our FAQ.

And if you have ideas for where Honoramma should evolve next, we’re collecting feature suggestions publicly, you can vote or suggest your own.

Building this openly matters to me.

It reminds me Sims or Happy Island on FB :D Good old times :D

Congrats! I was genuinely surprised (in a good way!) to see you mention me in your maker comment 😊

By the way, why does the park layout get locked after activation?

Love the concept 🙂 Can monuments be dedicated to anyone? For example, could I create a park for my dog?

Nice idea 🙂 How customizable are park layouts and themes?

Dear everyone who left questions, a few of you told me that I’m not replying and seem inactive.

That’s not the case. I’ve been here since the morning, as recommended for Product Hunt launches, ready to respond to every comment in real time. Unfortunately, I simply don’t see your comments on my side at the moment.

Thank you very much for your support and for your questions. I know they’re there, even if I can’t see them yet.

Interesting project 👍 How do you prevent abuse or inappropriate content?

Really curious about this. If someone buys a slot in my public park, do I earn anything from that?

Congrats on the launch! 🎉 This looks interesting. How is it different from something like MyHeritage or Geni?