Tooli is your AI planning assistant inside Toolio. It lives alongside your plan as a real, immersive chat — ask questions in plain English about Merchandise Plan, Assortment Plan, Item Plan, and Allocation, and get answers grounded in your live data. With Skills, your team can turn its repeatable workflows into reusable slash commands that anyone can run.

Open Tooli from the AI button in the top app bar. You can use it in two modes:

Switch between modes from the Tooli header.

Every conversation is saved automatically. Find your past threads in the History sidebar inside Tooli.

Tooli reads from the same modules your team uses every day, through the Toolio MCP:

Tooli sees the live state of your plan — no stale exports, no separate copies. Every answer is traceable back to the underlying slice of the plan.

Tooli is read-only on Merchandise Plan, Assortment Plan, and Item Plan. It can analyze your plan, surface variance drivers, and draft recommendations, but it won't change a number on those modules.

On Allocation, Tooli can take a small set of write actions — for example, updating an Allocation Strategy on your behalf. Tooli won't make any changes unless you explicitly ask it to. If you do, it will tell you exactly what it changed and link you back to the affected strategy so you can review.

Governed write-back across the rest of the platform is coming with Agent Hub.

Beyond reading data, Tooli can take a few supporting actions inside the conversation:

A Skill is a saved playbook your team writes once and runs any time, by typing a slash command in Tooli.

For example, a Monday-morning recap Skill might say:

Pull last week's sales vs. plan and vs. LY at the category level. Flag any category off-plan by more than three points. Break out top variance drivers from the assortment level. End with a one-paragraph narrative ready for the trading team's standup.

The first time your senior planner wrote that recap by hand, it took two hours. With a Skill, it's /monday-morning-recap — every time, same shape, same rigor, every category covered, live from the plan.

Skills come in three scopes:

A Skill's scope is determined by where it was created:

To run a Skill, type / in the Tooli input box. The skill picker opens with everything available to you — system, organization, and personal. Keep typing to filter, pick the Skill you want, and Tooli runs the workflow.

You'll see the Skill name as a pill in your message, so it's clear which playbook was used to produce the answer.

To author a Skill, go to Settings > Tooli > Skills. From there you can:

Skill bodies are written in Markdown, with a 32KB body limit.

The fastest way to write a Skill is to let Tooli draft it from a conversation you've already had. When you work through an analysis once in chat and like the result, you can capture that exact workflow as a Skill so it runs the same way every time — for any store, category, or week, and for anyone on your team. This is the recommended path when you find yourself wanting to "save this conversation" or "save this prompt to run again later."

Follow these steps:

The key is to pin down the decision junctures — the points where the workflow could reasonably go more than one way. A spec that says "look at slow-moving items" leaves room for interpretation, so two people can get two different answers. A Skill that says "flag any item with fewer than two units sold in the last eight weeks and more than four weeks of supply on hand" gives the same answer every time. Naming those choices explicitly in the Skill body is what makes the result repeatable across your team — and it's why a saved Skill is more reliable than a prompt you've pasted into a document.

A conversation in History is a record of what you asked; a Skill is a button you can press again. Promote the conversations you find yourself repeating into Skills so you never have to rebuild them by hand.

Toolio seeds your tenant with a starter library of built-in Skills you can use as-is or fork into a tenant-specific variant. They show up in your Skills list marked as System.

Open Settings from inside Tooli to:

Tooli and Skills are in public beta and included for every tenant where they're enabled at no additional cost while we refine the product.

At general availability, Tooli and Skills will move to a freemium model — included with every Toolio seat up to a generous usage threshold, with usage-based pricing beyond that. The same model will extend to Agent Hub when it ships. We'll give you clear advance notice before any pricing changes take effect.

If you don't see the AI button in the app bar, ask your account team about enabling Tooli for your tenant.

Conversations are kept for 30 days, with up to 100 conversations per user and 500 messages per conversation. After 30 days of inactivity, a conversation expires.

No. Conversations are private to your user. Sharing is not yet available.

It depends on the module. Tooli is read-only on Merchandise Plan, Assortment Plan, and Item Plan — it can analyze and recommend, but every change there still happens through you.

On Allocation, Tooli can take a small set of write actions (for example, updating an Allocation Strategy) — but only when you explicitly ask it to. It will not make any updates on its own, and it will tell you what it changed so you can review.

Governed write-back across the rest of the platform is coming with Agent Hub.

The Skill picker still opens — you'll see the built-in System Skills that Toolio ships with your tenant. You can run them as-is or fork them into an organization variant.

Create or save it under the Organization scope from Settings > Tooli > Skills. Anyone in your tenant with access to Tooli will see it in their Skill picker.

The model catalog depends on which providers your tenant has configured. Some tenants have only Anthropic Claude available; others have Anthropic plus Azure OpenAI.

Open Tooli Settings and click Reset assistant connection. That clears Tooli's session with the Toolio MCP and re-mints your connection on the next message.

Yes — Skill bodies are capped at 32KB. The editor shows you the size as you type. If you bump up against the limit, split the work into two Skills or trim the body.