Power Guides · Restamped 2026

Build an AI operating system. Then give it a harness.

Defaults beat prompts. Workspaces beat long threads. Verification beats hope. Start with the Trinity card if you do not know which model to open. Install Operator OS for documents. Install the Agentic Coding Harness for a repo.

Built for accuracy, leverage, and clean execution. Starts at €4.99.

Last verified: 2026-08-31

The lineup

One system. Four clear jobs.

Trinity Routing Card

€4.99

Which surface or model family fits this task.

Pick a surface in five minutes. Documents live in ChatGPT Projects. Repos live in a coding agent. Restamped quarterly so model names do not age inside the operating-system guides.

For
Everyone
Includes
Compass and chain prompts.
Not included
Install surfaces.

Operator OS

€29–39

ChatGPT as a workspace OS.

Set up files, project instructions that override global Custom Instructions inside the workspace, project-only or default memory, sharing side effects, and the Brief → Plan → two-pass → Verify loop.

For
PMs, analysts, writers, and operators
Includes
Projects, Memory policy, three Custom GPTs, Brief to Verify, and an asset zip.
Not included
Codex CLI and Claude Code.
Install the ChatGPT OS · TODO: gumroad slug

Agentic Coding Harness

€49–79

Codex and Claude Code as a repo OS.

Install instruction precedence, SKILL.md procedures, one MCP at a time, plan mode, secret-denying hooks, and a human gate for irreversible work. One captain per repo per session.

For
Developers and technical operators
Includes
AGENTS.md, CLAUDE.md, Skills, MCP, hooks, plan and verify, and an asset zip.
Not included
A ChatGPT Projects deep-dive.
Install the coding harness · TODO: gumroad slug

Anti-Robot Writing Guide

€9–19

Human voice on top of any draft.

Use the 2026 cringe list, anecdotal injection, and an optional Voice GPT to make correct drafts sound like a person with a point of view.

For
Creators and anyone publishing
Includes
Burstiness profile, cringe list, voice-memo injection, and Voice GPT instructions.
Not included
A second operating system.
Get the writing guide · TODO: gumroad slug

Operator Stack: Operator OS + Agentic Coding Harness + Anti-Robot + Trinity. €79–99. TODO: gumroad slug.

Comparison

Choose the surface before you buy the system.

Trinity Routing Card · €4.99

Price
€4.99
Job
Surface or model family
Buyer
Everyone
Install time
5 min
Includes
Compass and chain prompts
Does not include
Install surfaces

Operator OS · €29–39

Price
€29–39
Job
ChatGPT workspace OS
Buyer
Document operators
Install time
30 to 90 min
Includes
Projects, Memory, Custom GPTs
Does not include
Coding-agent install

Agentic Coding Harness · €49–79

Price
€49–79
Job
Codex and Claude Code repo OS
Buyer
Technical operators
Install time
45 min to 3 hr
Includes
Instruction files, Skills, MCP, hooks
Does not include
ChatGPT Projects deep-dive

Anti-Robot Writing Guide · €9–19

Price
€9–19
Job
Human voice
Buyer
People who publish
Install time
20 min
Includes
Voice system and cringe list
Does not include
A second OS

The workspace layer

Projects means this now.

  • ChatGPT Project: a workspace with files, project instructions, and a memory mode.
  • Project instructions override global Custom Instructions inside the workspace.
  • Choose default memory or project-only memory for the work at hand.
  • Use a Temporary Chat when you do not want the system to learn from the task.
  • The solution is not better prompts. It is a better system.

Migration

Already bought an earlier guide?

If you bought Lite or Pro, you already own the philosophy. Global Custom Instructions stay. The charter moves into a real Project file. You are grandfathered into Operator OS.

FAQ

The product boundary.

Why separate Operator OS and the Harness?

Documents and repos do not share an output contract. A ChatGPT workspace guide cannot teach AGENTS.md or hooks.

Will the free GPT-5 playbook replace Operator OS?

No. The playbook is a model map. Operator OS is the workspace install. Paid stays ahead of free, not behind it.

Free chapter

Layer-0

The free chapter is the next curriculum release. It will teach the mental model before an install.

Free tool

Prompt Scorecard

Grade a prompt before a model turns it into a confident mess.

Grade this prompt

Free primer

Agentic coding stack

See the current coding-agent comparison before you install a repo harness.

Read the primer

Verified buyer reviews

What buyers said.

“An excellent guide with checklists to make sure you give the most accurate prompt to get the most relevant result.”

Anonymous

“Perfect for beginners who finally decided to use AI in the most efficient way possible.”

Halil Önder Özbaşak