Skip to main content
CHU

Independent product studio

I build practical AI systems that turn real workflows into repeatable products.

Tested skill work packs, tools, and case studies for people building with coding agents.

Explore Products

Built by CHU KIT LUN

Independent product builder

Small products with a clear job, a visible scope, and a real way to start.

AI Operator Workflow — FREE

HK$0

Outcome: Turn one idea into a clear next step with a small, structured AI workflow.

1.0.0 · Codex / ChatGPT skill hostView product

AI Operator Workflow — PLUS

HK$59

Outcome: Turn ideas into validated, launch-ready products with AI, even if you do not code.

1.0.0 · Codex / ChatGPT skill hostAvailableView product

AI Operator Workflow — PRO

HK$159

Outcome: Build your personal AI operating system with advanced Codex execution, handoffs, verification, and compounding workflows.

1.0.0 · Codex / ChatGPT skill hostAvailableView product

Codex Starter Skill

Free

Outcome: A small Codex skill for turning a repeated instruction into a reusable workflow with clear triggers, constraints, and validation.

1.0.0 · CodexView product

SEO Autopilot — Codex Skill Work Pack

HK$199

Outcome: A structured Codex work pack for repeatable SEO research, prioritization, implementation, quality review, and reporting.

1.0.0 · CodexAvailableView product

Build → Use → Verify → Productize

CHU turns working practice into products only after the workflow has earned its shape.

  1. 01

    Build

    Start with a bounded real workflow.

  2. 02

    Use

    Put the system into ordinary work.

  3. 03

    Verify

    Check the result across languages and environments.

  4. 04

    Productize

    Package what is useful and name the limits.

Evidence, not theatre

A product is easier to trust when its scope and proof are visible.

  • Three language editions
  • Versioned release artifacts
  • Browser-verified workflows
  • Accessibility checked

Building Now

Experiments in progress, shown with their current state.

ModelPilot

Testing

A public beta that turns one pasted Codex task into a recommended model and reasoning effort.

Status
Testing
Problem
Users often do not know which Codex model and reasoning effort fit a task.
Current hypothesis
A single-paste transparent recommendation is useful enough to validate before building an automatic router.
Evidence collected
A working trilingual browser prototype and deterministic rule tests. No usage, accuracy, or time-saving measurement exists yet.
Next decision
Measure whether real users follow the recommendation and come back to use the tool again.
Last updated
2026-08-20
View research

CHU · Independent Product Studio

Build with a clearer workflow.

Explore the products, inspect the work, and follow the decisions behind them.

Explore Products