MIA — Business Site

A multi-page business site on Cloudflare Pages aimed at local small businesses, with a local development server that behaves like the live platform so routing bugs appear before launch.

Closed the gap between "works on my machine" and "works deployed"

Visit the live site

74,875lines of code
119files
102commits
4HTTP endpoints
Screenshot of MIA — Business Site

The hard part

Edge platforms rewrite addresses and route requests in ways an ordinary local server does not. So a whole class of bug is invisible until after you go live, and the feedback loop for fixing it runs through a deployment every time. The only real answer is to make local behave the way production does.

The development server reproduces the platform's own clean-URL rewriting and worker proxying, so what is tested locally is what actually ships. Static-first, with the dynamic surface isolated at the edge.

Local and edge routing kept identical — deploys gated by environment — accessibility handled structurally on the inner pages

What the measurements mean

Only what is unusual for this project. Every sentence names the field it came from; nothing here is asserted that cannot be checked against the data.

How it works

Edge platforms, which run a site on servers spread close to its visitors, rewrite addresses and route requests in ways an ordinary local server does not. A whole class of bug therefore stays invisible until the site is live, and every attempt to fix one goes through another deployment. The answer was to make local development behave the way production does. The development server reproduces the platform's own clean-URL rewriting and Worker proxying, so what is tested locally is what actually ships. The site is static first. Its dynamic parts, the chatbot and contact endpoints, sit in a Worker at the edge that proxies them to a FastAPI backend. Accessibility is handled in the structure, with skip links and keyboard navigation on the inner pages. Development and production deploys are separate. A production deploy asks for the word LIVE to be typed, and only public files are staged for publishing. Since October 2026 the home page speaks directly to local small businesses, with an invoice demo and a payback calculator whose examples are marked as fictional.

What kind of engineering this is

Six signals counted from the source: files that run models locally, that act without being asked each time, that check their own output, that survive failure, that touch parsers and protocols, and that were measured then improved. Every axis carries its file count.

Depth profileDepth signals: Local inference 3 files, Autonomy 0 files, Verification 3 files, Resilience 19 files, Low level 5 files, Iteration 0 filesLocal inference3 filesAutonomy0 filesVerification3 filesResilience19 filesLow level5 filesIteration0 files

The API, by shape

Routes read out of the source by a parser and grouped by their first path segment. They come from the code rather than the README, because a README can fall behind the code it describes.

HTTP surface2 routes across 2 groups: //, /<path:filename>//1 routeROUTE//<path:filename>1 routeROUTE/<path:filename>

What it is written in

Share of code bytes. Vendored bundles and generated output are set aside before this is computed, which is why it differs from a simple file count.

Language mixHTML 80.9%, CSS 7.3%, JavaScript 6.9%, Python 3.8%, Batchfile 0.6%, PowerShell 0.4%HTML 80.9%

What stands out

Capabilities

Grade

Every check names the field it measured. Checks that cannot apply are excluded, never counted as failures.

A
94.1% over 18 of 18 checks100% coverage

Not met

Architecture

Directories carrying real modules, largest first.

HTTP surface

4 endpoints in total.

Vulnerable dependencies

12 declared dependencies match a published advisory. Checked offline against the OSV database; only pinned versions can be checked, so anything unpinned here is unknown rather than clean.

Vocabulary

Parsed 14 of 14 Python file(s).

70functions
11methods
1class
10documented modules

Classes

Functions

Depth

Signals counted from the code, not asserted.

102 commits across 7 separate working days over 11.5 months, 49 of them revising existing behaviour

Built from scratch

Proves itself

Survives failure

Local inference

Systems

Environment engineering

Working within limits

Networking

Distributed systems

Operating-system internals

Languages

Built with

Recurring themes in the commit history

Evidence

Taken from the code itself: a module docstring or the opening of the README.

MIA Intake Backend Serves chatbot (via LM Studio) and contact form submissions. Runs on port 8001, exposed via Cloudflare Tunnel as resume.rnsu.net.

At a glance

Status
maintained
Primary language
HTML
First commit
2025-09-16
Last activity
2026-08-31
Domain
Automation & scraping
Agent docs
CLAUDE.md, AGENTS.md, PROJECT_KNOWLEDGE.md, .claude, .superpowers