Show HN за 18 октября 2025 г.

My dad just bought a new printer. I joked and asked if it also had fax and a scanner. He said yes, not realizing I was making fun of him, so I asked if it also supported Morse code. That made it clear :) Anyway, I decided to take the joke too far and build a silly Morse code chat app for fun.

ServiceRadar is an open-source platform for distributed, secure network management and observability, scaling to 100k+ devices. Born from frustration with complex traditional NMS tools like Zabbix, it bridges legacy (SNMP/syslog) and modern (gNMI, OTLP) protocols for cloud-native environments.

We built ServiceRadar to simplify monitoring hybrid telecom networks, evolving it into a Kubernetes-native solution with Helm and Docker support. It uses mTLS with SPIFFE/SPIRE, NATS JetStream for event streaming (90M+ EPS), and SRQL for intuitive queries. Integrated with OpenTelemetry, Prometheus, and CloudEvents, it fills the network visibility gap in CNCF’s application-focused observability stack.

We’re seeking early adopters to try our demo or deploy locally—no sign-up needed. Feedback on usability or contributions for new protocols would be awesome.

Quick Start: helm install serviceradar carverauto/serviceradar or docker compose up -d

GitHub: https://github.com/carverauto/serviceradar (please star!)

Docs: https://docs.serviceradar.cloud

Join our Discord or use GitHub Issues to share thoughts.

We have another election coming up at the end of this month here in the Netherlands, and how we use our limited land is a hotly debated topic.

Last year also I shared this animated hexagonal map of Dutch land use. I've now expanded this to include several more European countries, so you can see what makes Dutch land use so special - or not.

If you want to help add more countries, I'd appreciate PRs on GitHub. The map rendering is already there, it only requires land use data which can be found in the SQLite database (at least for EU countries).

Github: https://github.com/vnglst/onsland Previous HN: https://news.ycombinator.com/item?id=40599763

We implemented Stanford's Agentic Context Engineering paper which shows agents can improve their performance just by evolving their own context.

How it works: Agents execute tasks, reflect on what worked/failed, and curate a "playbook" of strategies. All from execution feedback - no training data needed.

Happy to answer questions about the implementation or the research!

Moin,

to learn Rust I decided to create a simple clone of the original lunar lander game. I would love to hear feedback on the quality of the code!

Building interactive widgets for ChatGPT turned out to be way more complex than I expected. Between the MCP protocol, widget registration, build setup, CSP policies, and resource serving — I kept writing the same boilerplate over and over just to get something basic working.

That’s why I built FastApps, an open-source framework that lets you build ChatGPT apps with only two files: a Python backend and a React frontend.

It removes all the glue code and focuses on developer experience:

- Auto-discovers and registers MCP widgets automatically

- Builds and bundles React UIs with zero config

- Manages CSP, metadata, and resource serving out of the box

- Type-safe data flow with Pydantic and React hooks

- Runs as an MCP server, fully compatible with ChatGPT and other MCP clients

If you’ve been experimenting with MCP or building ChatGPT tools, I’d love your feedback. It’s fully open source — and I’m looking for contributors who want to help shape this ecosystem together.

GitHub → https://github.com/DooiLabs/FastApps Docs → https://www.fastapps.org/

Bottleneck is a desktop app for code reviewing AI-written pull requests.

The goal is to be able to track PRs on GitHub authored by agents (i.e. Codex, Devin, Cursor, Claude Code) and compare branches. So if you throw multiple coding agents at a ticket, this would be an easier way to let agents "bake off" against each other and pick the best one. (No need to open the Github website and switch between slow loading tabs).

I'd love feedback from any power users who are deep with AI coding agents.