// live · dubai, uae
I'm Suhail Mohebi. I build every day with a small fleet of AI coding agents — Claude, Codex, Droid and Cursor — mostly by voice. This site is our open lab notebook: the tools, settings, recipes and methods behind the human × AI partnership, logged in public.
CHOOSE BY INTENT
Pick the path that matches what you need now. Each destination has one clear job.
A small, repeatable system turns a vague request into work that can be checked, trusted, and continued.
Read the source of truth and the constraints before touching anything.
Define the outcome, scope, evidence, and stop conditions.
Make the smallest coherent change that solves the real problem.
Test behaviour, content, performance, links, and responsive layout.
Leave a concise state packet so the next run starts from evidence.
ORIENTATION BEFORE ACTION
The folder nearest the agent may not own the page people are using. Trace the public result back to its exact editable source before changing it.
Newest first. Filter by thread, or hit ⌘K to search the Notebook.
Showing all 15 notebook entries.
The notebook you're reading was designed, coded, and deployed inside a single conversation. I pointed at a domain that had sat untouched on a website builder for years; the AI checked its DNS, proposed the migration, and built this — aurora gradients, particle field, command palette and all. Very meta. Hello, world.
READ THE METHOD See how a rough brief becomes a connected, published page.
We planned, wrote, cross-linked, and published a full topical content cluster — hub pages, reference guides, and spokes — in one overnight session, complete with citations, structured data, and social cards generated to match the template family. Proof that an AI can hold a whole content architecture in its head at once.
READ THE METHOD Follow the page-production workflow behind this overnight content run.
Instead of one AI reading one file at a time, we fan out: a dozen specialised agents audit a whole site in parallel — accessibility, structured data, performance, links, consistency — then a separate adversarial council tries to refute every finding before it reaches me. False positives die in review, not in production.
READ THE METHOD Use the evidence ladder behind an adversarial site review.
A full visual redesign of one page template rolled out across twenty pages — new hierarchy, cleaner layout, consistent design tokens — verified page by page in a real browser before shipping. The AI takes the screenshots; the human takes the decisions.
READ THE METHOD See the verification loop used to check a rollout page by page.
Search is splitting: classic blue links on one side, AI answers on the other. We build entity hub pages, definition-first answer blocks, and machine-readable facts so primary sources are easy for search engines and AI assistants to retrieve and cite. SEO's next chapter, tested in public.
We'd banned a certain inline prompt for looking pushy. Then analytics showed pages carrying a subtle, dismissible version retained readers far better than we assumed — so we reversed the rule, rolled it out only where relevant, and pinned the list with a test so it can't silently grow.
READ THE METHOD Turn a design hunch into a falsifiable, evidence-led decision.
We shipped a site chatbot built around a curated knowledge base and a live-fetched fact source. It answers visitor questions, proposes actions the server validates before running, and can hand a conversation to a human. Built, tested, and iterated across several sessions of this partnership.
Mobile readers don't scroll through essays. We restructured a mobile template around a lean above-the-fold — the essentials first, the deep detail collapsed behind intent. Load less, keep more.
The most useful pattern we've found: don't let one AI pass-grade its own homework. Significant changes go through sign-off batches — independent review lenses, contamination sweeps for stale branches, and pre-push audits that catch what a single context window can't hold.
READ THE METHOD Open the independent-review recipe behind this sign-off pattern.
One change can touch dozens of surfaces: the page, the indexes, structured data, sitemaps, machine-readable files, tests. We turned that into a repeatable checklist the AI runs end to end — so a launch is a review, not a scramble.
READ THE METHOD Use the blast-radius and verification recipes behind this launch checklist.
The library passed eighty published articles — every one fact-checked against primary sources, structured for both human readers and machine retrieval, and run through an automated ten-point checklist before it ships.
READ THE METHOD Follow the publishing workflow that turns repeated pages into a durable system.
We split one project into two sites with distinct roles — a primary destination and a separate evidence-first layer for references and analysis — deliberately connected but never cross-wired. Two jobs, cleanly separated.
An interactive calculator that turns confusing unit math into a few guided inputs — built mobile-first with progressive disclosure and shipped after real-device testing. Still one of the most-used tools we've built.
On 28 February 2026, Iran launched missile and drone attacks on the UAE. A Shahed-238 drone struck the forecourt at Fairmont The Palm, while air defences intercepted incoming weapons over Dubai (UAE Ministry of Foreign Affairs; Human Rights Watch). We were lucky; many weren't. That week changed how I work: everything that could be delegated got delegated, verified, and shipped. Every entry above this one traces back to that decision.
A founder in Dubai, a static site, and an AI in a terminal. What began as "help me fix this page" became a standing partnership with its own memory, playbooks, and division of labour. This notebook is the public record of where it goes.
READ THE METHOD See how one AI partnership became a four-agent toolbench.
— beginning of the notebook —
The newest practical guides to workflows, tools, verification, context, recipes and reusable templates.
the human
Emirati, born and based in Dubai — grandson of Baqer Mohebi, whose 1931 shop on the Creek grew into one of the city's oldest family enterprises. Spent years as a derivatives trader; now an agentic engineer working hands-on with AI every day. Sets direction, makes the calls, owns the outcomes — and logs the craft here so others can borrow what works.
the AI fleet
Four AI coding agents, each best at something different — led by Claude, with Codex, Droid and Cursor alongside, plus local models for private work. They write the code, run the audits, draft the research, and keep the receipts. Everything here was built in conversation — prompt by prompt, commit by commit.
Most human–AI work happens in private chat windows and disappears. We think the interesting part is the collaboration itself — what gets delegated, what gets verified, where the AI shines and where the human has to steer. So we log it here, in the open, as it happens.
Find Notebook entries, Deep Dives, and copy-ready templates.