/portfolio/AI-page-generator

AI Page Generator

A local-LLM-assisted one-shot page maker: prompt in, HTML out, then serve the generated file with as little sharing friction as possible.

back to index -> read project notes

projectAI Page Generator

statusrapid publishing tool

surfaceLLM web generation

index06 / 06

stack tagslocal LLM, HTML/CSS/JS, prompting

local LLMHTML/CSS/JSpromptingserver routingstatic file serving

01 / project notes

What it was actually solving.

AI Page Generator is a small publishing utility for making quick one-file web pages and sharing them without forcing the recipient through setup or tooling.

Challenge

The hard part was not only getting an LLM to produce a page. It was stripping away the extra commentary models love to add, saving clean code, and serving the result reliably from Calvin’s own server.

Solution

A generation flow with a stricter system prompt, cleanup logic that extracts the usable HTML, and a server route that can publish generated files under the view path for easy sharing.

What stuck afterward

This pushed the boundary between “toy generator” and “useful web tool.” It taught more about exposing server functionality safely while keeping the user path simple: make page, publish page, send link.

02 / video demo

Proof it actually moved.

A quick demo clip from the original project page, kept inside the same hard-edged terminal frame instead of burying it below the writeup.

demo source/static/portfolio/demo-videos/generatorpagedemo.mp4

03 / keep digging

More things from the same lab bench.

Same visual system as the main ToadTech revamp: hard-edged, terminal-ish, and allergic to generic portfolio gloss.

LUCID DashboardControl panel for the home network: start machines, run scripts, launch services, and avoid SSHing into six different boxes for one simple action.CVMUN Conference WebsiteA public site for Champlain Valley Model UN that puts schedules, committees, guides, rules, and contact info where delegates can actually find them.Portable OllamaA zip-based local LLM setup that can run from a folder or USB drive without asking the host machine for a whole installation ritual.