What is an AI playbook?
Most people use AI assistants the same way: open a chat, type a request, get a draft, fix the draft. Next week brings the same task, the same explanation, the same fixes. The assistant never remembers. An AI playbook exists to break that loop. It is a saved, reusable definition of how you want a specific kind of writing done — built once, improved over time, and applied automatically every time you call it.
What does an AI playbook store?
A playbook holds four kinds of information.
Tone. How you sound. Direct or warm. Formal or casual. Whether you open with the point or build up to it.
Rules. Hard constraints the assistant must follow every time. Never use buzzwords. Keep emails under 150 words. Always end with a clear next step.
Examples. Real samples of your writing. Examples carry information that instructions cannot. Three good emails teach rhythm and word choice better than a paragraph of description ever will.
Learned preferences. This is the part that separates a playbook from a static instruction set. When you correct an output — shorter, less formal, drop the filler — the correction is saved as a preference. The playbook becomes more accurate with use, the way a new colleague does.
A playbook is usually scoped to one role or one job: sales outreach, manager updates, customer support replies. Narrow scope is a feature, not a limitation. A playbook that tries to cover everything ends up describing no one.
Why does calling it by name matter?
A playbook has a name, and the name changes how you use it. You do not open an app, paste context, and explain yourself from scratch. You invoke the playbook where the text already is.
Highlight a rough draft and send it to the playbook from the share sheet. Ask Siri to write the reply using your Sales playbook where supported. Trigger it from a shortcut in the middle of other work. The playbook stops feeling like a tool you operate and starts feeling like a reusable writing behavior you can call by name.
Naming also keeps roles separate. Sales and Manager are different voices with different rules. Calling one by name guarantees you get that voice, not a blend of everything you have ever asked for.
How is a playbook different from a prompt?
In one sentence: a prompt is an instruction you repeat, while a playbook is behavior that accumulates. Prompts reset to zero in every new chat; a playbook carries everything it has learned into the next draft. The full comparison, including the cases where a plain prompt is honestly the better tool, deserves its own page.
What does a playbook look like in ILURA?
The app is built around this exact concept. You create a playbook per role — Sales, Manager, Founder, or anything you define yourself. You train it by correcting its outputs: each correction becomes a learned preference, and preferences are versioned, so you can roll back if a change makes drafts worse. You invoke playbooks by name through Siri, the share sheet, or Shortcuts.
Generation happens on your iPhone with Apple Intelligence. There is no account to create, and the App Store privacy label reads Data Not Collected — the playbook, and everything it has learned about how you write, stays on the device. The free version includes three active playbooks, which covers most people’s actual roles.
How do you build a playbook worth keeping?
Start with one role, the one where you write the most. Give it a short tone description and two or three real examples of writing you were happy with. That is enough to begin.
Then use it on real work, and correct what comes back instead of silently fixing it. Silent edits teach the playbook nothing. Corrections become preferences, and preferences compound.
Review what it has learned every few weeks. Some preferences age badly — a rule that made sense for one difficult client can creep into everything. Prune those.
Expect the first week to be rough and the fourth week to be quiet. A good playbook is boring in the best way: drafts come back close to final, and you stop explaining yourself to software you already taught.
Turn the playbook into agent behavior
A playbook becomes more powerful when it is trained by correction. Each saved preference moves it from prompt text toward a private role agent.
- Start with one role
- Correct one real output
- Save the preference as readable behavior
Try it now
Put this to work on a real message.
Open ILURA, bring in a message you actually need to handle today, and get it done in your voice — free, on device, no account. It learns the preference, so the behavior carries to the next one.
Free to start · No account · Data Not CollectedQuick answers
- Is a playbook the same as a prompt?
- No. A prompt is an instruction you repeat in every chat. A playbook is saved behavior that accumulates corrections over time and is invoked by name.
- Do I need a technical background to build a playbook?
- No. You build one by describing how you write, adding a few real examples, and then correcting outputs you dislike. Each correction becomes a saved preference.
- Where does an ILURA playbook live?
- On your iPhone. Generation runs on the device with Apple Intelligence, there is no account, and the App Store privacy label reads Data Not Collected.