John Vedovino PRO

I built a matchup win-rate board for One Piece TCG based on 132 real simulator games. The data is from a single player seat (Sabo v3, OP18 meta) - not a community survey, but real logged game outcomes with confidence intervals. The board shows win rates only when the sample is large enough to be statistically meaningful (n>=5, Wilson lower bound). Below that threshold it shows UNMEASURED rather than a misleading number. Three matchup cells are measured: vs Mihawk 67%, vs Nico Robin 60%, and the Sabo mirror at 50%. The useful part for anyone tracking deck performance: log game results, let the math tell you when a verdict is real vs noise. One season of games is enough to get 3-5 measured matchups. — John (JohnVeds15). I build tools with Claude Code; some of this came out of tracking my own games.https://huggingface.co/spaces/JohnVeds15/optcg-matchup-board

posted an update 2 days ago

Post

108

I built a shared Claude assistant for families that keeps memories across context resets.

The problem: after weeks of conversation, the context window fills. Old turns get cut. You lose the schedules, preferences and shared history that made the assistant actually useful — there's no clean way to carry any of it forward.

Silta solves this with a persistent memory bridge. Every family member talks to one shared assistant. Key facts — schedules, recurring preferences, decisions — get extracted automatically. When context fills, old turns compress to summaries. Nothing load-bearing is lost.

The demo at the Space lets you type to the assistant and watch the memory panel update in real time, including what would survive a context reset.

— John (JohnVeds15). I build agent tooling with Claude Code; some of this came out of my own scripts.https://huggingface.co/spaces/JohnVeds15/silta-family-memory

1 reply

posted an update 8 days ago

Post

Built a pure-browser OSM flyover demo yesterday — no backend, just OpenStreetMap tiles rendered in three.js.

Pick any of 12 preset cities (New York, Tokyo, London, Paris...) and a camera flies a looping path over the actual street grid at WebGL speed. The tile pool only loads what's in view, so memory stays bounded while the flight continues.

The hard part was projecting tile coordinates into three.js world space without seams. Once that was sorted, adding a smooth flight path was straightforward. Everything runs in a single HTML file — the only external calls are the OSM tile server and three.js from CDN.

Demo: JohnVeds15/fly-over-real-cities

— John (JohnVeds15). I build agent tooling with Claude Code; some of this came out of my own scripts.https://huggingface.co/spaces/JohnVeds15/fly-over-real-cities

posted an update 16 days ago

Post

I built a small agent that answers Hugging Face threads, and questions on my own datasets, on a schedule, without me approving each one.

It runs four times a day. Each run it can post at most twice, capped at five total per day, and it tracks how many are queued versus already sent. Posts to the Hub go through the Hub API with a fine-grained token; if that comes back 401 or 403, the agent queues the post instead of failing. Forum replies work the same way: without a Discourse API key, they get written to a local queue file instead of erroring out.

The lesson I'd pass on: build the failure path before the happy path. An autonomous posting agent hits auth errors and missing keys more than you'd expect, and the difference between queuing and crashing is what makes it safe to leave running unattended.

— John (JohnVeds15). I build agent tooling with Claude Code; some of this came out of my own scripts.

John Vedovino PRO

AI & ML interests

Recent Activity

Organizations

JohnVeds15's activity