The best Hacker News stories from Show from the past day
Latest posts:
Show HN: I made an app that lets you save audio to your Cameral Roll
...because Apple make it too hard to navigate chronologically through anything but your photos. Really they should have let you do this directly from voice memos so that you can share whenever and to wherever you want.<p>If anyone has any suggestions to make it better, let me know. Thanks.
Show HN: EnrichMCP – A Python ORM for Agents
I've been working with the Featureform team on their new open-source project, [EnrichMCP][1], a Python ORM framework that helps AI agents understand and interact with your data in a structured, semantic way.<p>EnrichMCP is built on top of [MCP][2] and acts like an ORM, but for agents instead of humans. You define your data model using SQLAlchemy, APIs, or custom logic, and EnrichMCP turns it into a type-safe, introspectable interface that agents can discover, traverse, and invoke.<p>It auto-generates tools from your models, validates all I/O with Pydantic, handles relationships, and supports schema discovery. Agents can go from user → orders → product naturally, just like a developer navigating an ORM.<p>We use this internally to let agents query production systems, call APIs, apply business logic, and even integrate ML models. It works out of the box with SQLAlchemy and is easy to extend to any data source.<p>If you're building agentic systems or anything AI-native, I'd love your feedback. Code and docs are here: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a>. Happy to answer any questions.<p>[1]: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a><p>[2]: <a href="https://modelcontextprotocol.io/introduction" rel="nofollow">https://modelcontextprotocol.io/introduction</a>
Show HN: EnrichMCP – A Python ORM for Agents
I've been working with the Featureform team on their new open-source project, [EnrichMCP][1], a Python ORM framework that helps AI agents understand and interact with your data in a structured, semantic way.<p>EnrichMCP is built on top of [MCP][2] and acts like an ORM, but for agents instead of humans. You define your data model using SQLAlchemy, APIs, or custom logic, and EnrichMCP turns it into a type-safe, introspectable interface that agents can discover, traverse, and invoke.<p>It auto-generates tools from your models, validates all I/O with Pydantic, handles relationships, and supports schema discovery. Agents can go from user → orders → product naturally, just like a developer navigating an ORM.<p>We use this internally to let agents query production systems, call APIs, apply business logic, and even integrate ML models. It works out of the box with SQLAlchemy and is easy to extend to any data source.<p>If you're building agentic systems or anything AI-native, I'd love your feedback. Code and docs are here: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a>. Happy to answer any questions.<p>[1]: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a><p>[2]: <a href="https://modelcontextprotocol.io/introduction" rel="nofollow">https://modelcontextprotocol.io/introduction</a>
Show HN: EnrichMCP – A Python ORM for Agents
I've been working with the Featureform team on their new open-source project, [EnrichMCP][1], a Python ORM framework that helps AI agents understand and interact with your data in a structured, semantic way.<p>EnrichMCP is built on top of [MCP][2] and acts like an ORM, but for agents instead of humans. You define your data model using SQLAlchemy, APIs, or custom logic, and EnrichMCP turns it into a type-safe, introspectable interface that agents can discover, traverse, and invoke.<p>It auto-generates tools from your models, validates all I/O with Pydantic, handles relationships, and supports schema discovery. Agents can go from user → orders → product naturally, just like a developer navigating an ORM.<p>We use this internally to let agents query production systems, call APIs, apply business logic, and even integrate ML models. It works out of the box with SQLAlchemy and is easy to extend to any data source.<p>If you're building agentic systems or anything AI-native, I'd love your feedback. Code and docs are here: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a>. Happy to answer any questions.<p>[1]: <a href="https://github.com/featureform/enrichmcp">https://github.com/featureform/enrichmcp</a><p>[2]: <a href="https://modelcontextprotocol.io/introduction" rel="nofollow">https://modelcontextprotocol.io/introduction</a>
Show HN: A DOS-like hobby OS written in Rust and x86 assembly
To try it out, simply build the project yourself from source, or use attached bootable ISO image of the system (in Releases on Github) and run it in QEMU.<p><a href="https://blog.vxn.dev/rou2exos-rusted-edition" rel="nofollow">https://blog.vxn.dev/rou2exos-rusted-edition</a>
Show HN: A DOS-like hobby OS written in Rust and x86 assembly
To try it out, simply build the project yourself from source, or use attached bootable ISO image of the system (in Releases on Github) and run it in QEMU.<p><a href="https://blog.vxn.dev/rou2exos-rusted-edition" rel="nofollow">https://blog.vxn.dev/rou2exos-rusted-edition</a>
Show HN: Claude Code Usage Monitor – real-time tracker to dodge usage cut-offs
I kept slamming into Claude Code limits mid-session and couldn’t find a quick way to see how close I was getting, so I hacked together a tiny local tracker.<p>Streams your prompt + completion usage in real time<p>Predicts whether you’ll hit the cap before the session ends<p>Runs 100 % locally (no auth, no server)<p>Presets for Pro, Max × 5, Max × 20 — tweak a JSON if your plan’s different<p>GitHub: <a href="https://github.com/Maciek-roboblog/Claude-Code-Usage-Monitor">https://github.com/Maciek-roboblog/Claude-Code-Usage-Monitor</a><p>It’s already spared me a few “why did my run just stop?” moments, but it’s still rough around the edges. Feedback, bug reports, and PRs welcome!
Show HN: Claude Code Usage Monitor – real-time tracker to dodge usage cut-offs
I kept slamming into Claude Code limits mid-session and couldn’t find a quick way to see how close I was getting, so I hacked together a tiny local tracker.<p>Streams your prompt + completion usage in real time<p>Predicts whether you’ll hit the cap before the session ends<p>Runs 100 % locally (no auth, no server)<p>Presets for Pro, Max × 5, Max × 20 — tweak a JSON if your plan’s different<p>GitHub: <a href="https://github.com/Maciek-roboblog/Claude-Code-Usage-Monitor">https://github.com/Maciek-roboblog/Claude-Code-Usage-Monitor</a><p>It’s already spared me a few “why did my run just stop?” moments, but it’s still rough around the edges. Feedback, bug reports, and PRs welcome!
Show HN: I wrote a new BitTorrent tracker in Elixir
Hello everyone!<p>I'm currently in a journey to learn and improve my Elixir and Go skills (my daily job uses C++) and looking through my backlog for projects to take on I decided Elixir is the perfect language to write a highly-parallel BitTorrent tracker.
So I have spent my free time these last 3 months writing one! Now I think it has enough features to present it to the world (and a docker image to give it a quick try).<p>I know some people see trackers as relics of the past now that DHT and PEX are common but I think they still serve a purpose in today's Internet (purely talking about public trackers). That said there is not a lot going on in terms of new developments since everyone just throws opentracker in a vps a calls it a day (honorable exceptions: aquatic and torrust).<p>I plan to continue development for the foreseeable future and add some (optional) esoteric features along the way so if anyone currently operates a tracker please give a try and enjoy the lack of crashes.<p>note: only swarm_printout.ex has been vibe coded, the rest has all been written by hand.
Show HN: I wrote a new BitTorrent tracker in Elixir
Hello everyone!<p>I'm currently in a journey to learn and improve my Elixir and Go skills (my daily job uses C++) and looking through my backlog for projects to take on I decided Elixir is the perfect language to write a highly-parallel BitTorrent tracker.
So I have spent my free time these last 3 months writing one! Now I think it has enough features to present it to the world (and a docker image to give it a quick try).<p>I know some people see trackers as relics of the past now that DHT and PEX are common but I think they still serve a purpose in today's Internet (purely talking about public trackers). That said there is not a lot going on in terms of new developments since everyone just throws opentracker in a vps a calls it a day (honorable exceptions: aquatic and torrust).<p>I plan to continue development for the foreseeable future and add some (optional) esoteric features along the way so if anyone currently operates a tracker please give a try and enjoy the lack of crashes.<p>note: only swarm_printout.ex has been vibe coded, the rest has all been written by hand.
Show HN: I wrote a new BitTorrent tracker in Elixir
Hello everyone!<p>I'm currently in a journey to learn and improve my Elixir and Go skills (my daily job uses C++) and looking through my backlog for projects to take on I decided Elixir is the perfect language to write a highly-parallel BitTorrent tracker.
So I have spent my free time these last 3 months writing one! Now I think it has enough features to present it to the world (and a docker image to give it a quick try).<p>I know some people see trackers as relics of the past now that DHT and PEX are common but I think they still serve a purpose in today's Internet (purely talking about public trackers). That said there is not a lot going on in terms of new developments since everyone just throws opentracker in a vps a calls it a day (honorable exceptions: aquatic and torrust).<p>I plan to continue development for the foreseeable future and add some (optional) esoteric features along the way so if anyone currently operates a tracker please give a try and enjoy the lack of crashes.<p>note: only swarm_printout.ex has been vibe coded, the rest has all been written by hand.
Show HN: Unregistry – “docker push” directly to servers without a registry
I got tired of the push-to-registry/pull-from-registry dance every time I needed to deploy a Docker image.<p>In certain cases, using a full-fledged external (or even local) registry is annoying overhead. And if you think about it, there's already a form of registry present on any of your Docker-enabled hosts — the Docker's own image storage.<p>So I built Unregistry [1] that exposes Docker's (containerd) image storage through a standard registry API. It adds a `docker pussh` command that pushes images directly to remote Docker daemons over SSH. It transfers only the missing layers, making it fast and efficient.<p><pre><code> docker pussh myapp:latest user@server
</code></pre>
Under the hood, it starts a temporary unregistry container on the remote host, pushes to it through an SSH tunnel, and cleans up when done.<p>I've built it as a byproduct while working on Uncloud [2], a tool for deploying containers across a network of Docker hosts, and figured it'd be useful as a standalone project.<p>Would love to hear your thoughts and use cases!<p>[1]: <a href="https://github.com/psviderski/unregistry">https://github.com/psviderski/unregistry</a><p>[2]: <a href="https://github.com/psviderski/uncloud">https://github.com/psviderski/uncloud</a>
Show HN: Unregistry – “docker push” directly to servers without a registry
I got tired of the push-to-registry/pull-from-registry dance every time I needed to deploy a Docker image.<p>In certain cases, using a full-fledged external (or even local) registry is annoying overhead. And if you think about it, there's already a form of registry present on any of your Docker-enabled hosts — the Docker's own image storage.<p>So I built Unregistry [1] that exposes Docker's (containerd) image storage through a standard registry API. It adds a `docker pussh` command that pushes images directly to remote Docker daemons over SSH. It transfers only the missing layers, making it fast and efficient.<p><pre><code> docker pussh myapp:latest user@server
</code></pre>
Under the hood, it starts a temporary unregistry container on the remote host, pushes to it through an SSH tunnel, and cleans up when done.<p>I've built it as a byproduct while working on Uncloud [2], a tool for deploying containers across a network of Docker hosts, and figured it'd be useful as a standalone project.<p>Would love to hear your thoughts and use cases!<p>[1]: <a href="https://github.com/psviderski/unregistry">https://github.com/psviderski/unregistry</a><p>[2]: <a href="https://github.com/psviderski/uncloud">https://github.com/psviderski/uncloud</a>
Show HN: Unregistry – “docker push” directly to servers without a registry
I got tired of the push-to-registry/pull-from-registry dance every time I needed to deploy a Docker image.<p>In certain cases, using a full-fledged external (or even local) registry is annoying overhead. And if you think about it, there's already a form of registry present on any of your Docker-enabled hosts — the Docker's own image storage.<p>So I built Unregistry [1] that exposes Docker's (containerd) image storage through a standard registry API. It adds a `docker pussh` command that pushes images directly to remote Docker daemons over SSH. It transfers only the missing layers, making it fast and efficient.<p><pre><code> docker pussh myapp:latest user@server
</code></pre>
Under the hood, it starts a temporary unregistry container on the remote host, pushes to it through an SSH tunnel, and cleans up when done.<p>I've built it as a byproduct while working on Uncloud [2], a tool for deploying containers across a network of Docker hosts, and figured it'd be useful as a standalone project.<p>Would love to hear your thoughts and use cases!<p>[1]: <a href="https://github.com/psviderski/unregistry">https://github.com/psviderski/unregistry</a><p>[2]: <a href="https://github.com/psviderski/uncloud">https://github.com/psviderski/uncloud</a>
Show HN: Free local security checks for AI coding in VSCode, Cursor and Windsurf
Hi HN!<p>We just launched Codacy Guardrails, an IDE extension with a CLI for code analysis and MCP server that enforces security & quality rules on AI-generated code in real-time. It hooks into AI coding assistants (like VS Code Agent Mode, Cursor, Windsurf), silently scanning and fixing AI-suggested code that has vulnerabilities or violates your coding standards, while the code it’s being generated.<p>We built this because coding agents can be a double-edged sword. They do boost productivity, but can easily introduce insecure or non-compliant code. One recent research team at NYU found that 40% of Copilot’s outputs were buggy or exploitable [1]. Other surveys mention that people are spending more time debugging AI-generated code [2].<p>That's why we created “guardrails” to catch security problems early.<p>Codacy Guardrails uses a collection of open-source static analyzers (like Semgrep and Trivy) to scan the AI’s output against 2000+ rules. We currently support JavaScript/TypeScript, Python, and Java, focusing on things like OWASP Top 10 vulns, hardcoded secrets, dependency checks, code complexity and styling violations, and you can customize the rules to match your project’s needs. We're not using any AI models, it's “classic” static code analysis working alongside your AI assistant.<p>Here’s a quick demo: <a href="https://youtu.be/pB02u0ntQpM" rel="nofollow">https://youtu.be/pB02u0ntQpM</a><p>The extension is free for all developers. (We do have paid plans for teams to apply rules centrally, but that’s not needed to use the extension and local code analysis with agents.)<p>Setup is pretty straightforward: Install the extension and enable Codacy’s CLI and MCP Server from the sidebar.<p>We’re eager to hear what the HN community thinks! Does this approach sound useful in your AI coding workflow? Have you encountered security issues from AI-generated code?<p>We hope Codacy Guardrails can make AI-assisted development a bit safer and more trustworthy. Thanks for reading!<p>Get extension: <a href="https://www.codacy.com/get-ide-extension" rel="nofollow">https://www.codacy.com/get-ide-extension</a>
Docs: <a href="https://docs.codacy.com/codacy-guardrails/codacy-guardrails-getting-started/" rel="nofollow">https://docs.codacy.com/codacy-guardrails/codacy-guardrails-...</a><p>Sources
[1]: NYU Research: <a href="https://www.researchgate.net/publication/388193053_Asleep_at_the_Keyboard_Assessing_the_Security_of_GitHub_Copilot's_Code_Contributions" rel="nofollow">https://www.researchgate.net/publication/388193053_Asleep_at...</a>
[2]: <a href="https://devops.com/survey-ai-tools-are-increasing-amount-of-bad-code-needing-to-be-fixed" rel="nofollow">https://devops.com/survey-ai-tools-are-increasing-amount-of-...</a>
Show HN: I built a tensor library from scratch in C++/CUDA
Hi HN,<p>Over the past few months, I've been building `dsc`, a tensor library from scratch in C++/CUDA. My main focus has been on getting the basics right, prioritizing a clean API, simplicity, and clear observability for running small LLMs locally.<p>The key features are:
- C++ core with CUDA support written from scratch.
- A familiar, PyTorch-like Python API.
- Runs real models: it's complete enough to load a model like Qwen from HuggingFace and run inference on both CUDA and CPU with a single line change[1].
- Simple, built-in observability for both Python and C++.<p>Next on the roadmap is adding BF16 support and then I'll be working on visualization for GPU workloads.<p>The project is still early and I would be incredibly grateful for any feedback, code reviews, or questions from the HN community!<p>GitHub Repo: <a href="https://github.com/nirw4nna/dsc">https://github.com/nirw4nna/dsc</a><p>[1]: <a href="https://github.com/nirw4nna/dsc/blob/main/examples/models/qwen2_5.py">https://github.com/nirw4nna/dsc/blob/main/examples/models/qw...</a>
Show HN: I built a tensor library from scratch in C++/CUDA
Hi HN,<p>Over the past few months, I've been building `dsc`, a tensor library from scratch in C++/CUDA. My main focus has been on getting the basics right, prioritizing a clean API, simplicity, and clear observability for running small LLMs locally.<p>The key features are:
- C++ core with CUDA support written from scratch.
- A familiar, PyTorch-like Python API.
- Runs real models: it's complete enough to load a model like Qwen from HuggingFace and run inference on both CUDA and CPU with a single line change[1].
- Simple, built-in observability for both Python and C++.<p>Next on the roadmap is adding BF16 support and then I'll be working on visualization for GPU workloads.<p>The project is still early and I would be incredibly grateful for any feedback, code reviews, or questions from the HN community!<p>GitHub Repo: <a href="https://github.com/nirw4nna/dsc">https://github.com/nirw4nna/dsc</a><p>[1]: <a href="https://github.com/nirw4nna/dsc/blob/main/examples/models/qwen2_5.py">https://github.com/nirw4nna/dsc/blob/main/examples/models/qw...</a>
Show HN: I built a tensor library from scratch in C++/CUDA
Hi HN,<p>Over the past few months, I've been building `dsc`, a tensor library from scratch in C++/CUDA. My main focus has been on getting the basics right, prioritizing a clean API, simplicity, and clear observability for running small LLMs locally.<p>The key features are:
- C++ core with CUDA support written from scratch.
- A familiar, PyTorch-like Python API.
- Runs real models: it's complete enough to load a model like Qwen from HuggingFace and run inference on both CUDA and CPU with a single line change[1].
- Simple, built-in observability for both Python and C++.<p>Next on the roadmap is adding BF16 support and then I'll be working on visualization for GPU workloads.<p>The project is still early and I would be incredibly grateful for any feedback, code reviews, or questions from the HN community!<p>GitHub Repo: <a href="https://github.com/nirw4nna/dsc">https://github.com/nirw4nna/dsc</a><p>[1]: <a href="https://github.com/nirw4nna/dsc/blob/main/examples/models/qwen2_5.py">https://github.com/nirw4nna/dsc/blob/main/examples/models/qw...</a>
Show HN: I made an online Unicode Cuneiform digital clock
I got the idea to pay homage to the people who gave us base-60 time, and in the process discovered that Cuneiform is in Unicode.
Show HN: I made an online Unicode Cuneiform digital clock
I got the idea to pay homage to the people who gave us base-60 time, and in the process discovered that Cuneiform is in Unicode.