Ask HN: What are tools you have made for yourself since the advent of AI?

I've made 3 that use pi as the primary interaction, with custom tools and scripts.

1. family tree based on wikitree format. Transcribe records, verify/edit, then incorporate them into the tree with full citations and biographies. This one is the big one. It includes a tree browser and best genealogical practices.

2. Pool Math replacement. Log pool chemistry tests to markdown files and suggest the right amount of chemicals to balance.

3. Calorie counter. Log calories to a markdown file, look up foods and amounts in online databases, sync with garmin connect for exercise calories.

All of these are written with AI but also are interacting with pi and telegram, mostly using deepseek v4 flash.

I made a golang socks5 proxy that routes traffic to different VPSes (or the default gateway) based on hostname, over mutual-TLS tunnels, authenticated using ed25519 keypairs shared out-of-band. The "client" and the "server" are the same piece of software, and there's a web UI for configuring the routes.

I made it to deal with internet censorship in the UK, where different sites have different optimal exit jurisdictions, and most sites work fine so I can avoid the extra hops where possible.

It also works well for video streaming sites with geofencing, since the geofence itself is usually implemented in the frontend rather than the CDN. So only the frontend traffic needs to be proxied, while the bulk CDN traffic doesn't need proxying at all.

Socks5 is the ideal layer for hostname-based routing, since the proxy can see DNS names without needing to sniff TLS SNI (which would be incompatible with ESNI/ECH etc.)

iirc it was basically all done in a single prompt, and I've been using it ever since. The only issue I've encountered is with WebRTC not working properly with some services. (Presumably it breaks the NAT holepunching process or something, I never diagnosed it)

Another project that isn't quite finished is a "universal" web video downloader that works by shimming the MSE APIs and remuxing the streamed segments back into a regular video file. The idea is that if you can watch it, you can save it - including but not limited to youtube videos. I started this one pre-AI but AI was a huge help with the container format wrangling.

[dead]

The overlap between ceramicists and technologists is never zero. Part of my initiative to slowly replace every part of my house and home with things I've made... 1% progress is good, right?

I wrote my own Claude Code agent. It leverages some parts of multiple different agent stacks but using my architecture. It's layered, has long-term memory, agency to expand upon work, access to linear, access to all the models and endpoints you could possibly wish, as well as support for combining multiple providers and models into the layers so that delegation to the workers happen on the cheaper, more cost effective models. All of this runs at scale in an orchestration platform I wrote, using said agent, to create a cluster of docker containers in a JBYOS configuration, cross-cloud, k8s or swarm or whatever can run docker. It's pretty sweet and is basically an AI software development shop. I only have to give it ideas and goals.