Skills, Behavior Studio & Discover

The Behavior Studio is where you shape how your Worker works: what it can do (skills), how much it acts on its own (initiative) and what it's allowed to do (rules). This article explains each concept, what happens "under the hood", and the discover mechanism that often surprises people at first.

The right picture: if the Worker is a teammate, the skills are its procedure sheets, the initiative is its temperament (wait vs take the lead), and discover is the moment it "pulls the right tool out of its box" when it needs it.

1. Skills

A skill is a reusable unit of know-how attached to the Worker. It combines:

• a title (e.g. Reply to an inbound support ticket)
• content (the procedure, the rules, examples — rich text)
• associated integrations (the tools useful to this skill)
• reference documents and links
• manager notes (context for the team)

How to create one

1. Open the Worker's page → Behavior Studio tab
2. Click Add a skill
3. Two modes:
   • Assisted mode (default): you describe the skill in natural language, the AI drafts it
   • Manual mode: you write the title and content yourself
4. Attach the relevant integrations / documents / links, then save (Cmd/Ctrl + Enter)

What happens under the hood: a skill is not code nor an executable tool. It's a knowledge document injected into the Worker's context. The Worker "reads" it when working in order to follow the right procedure.

Limit: up to 50 skills per Worker. A few clear skills beat fifty vague ones.

2. Initiative (the autonomy level)

The initiative defines how much the Worker acts on its own. Four levels:

You set this level with a slider at creation, then in the Profile tab (or Behavior Studio).

Under the hood: the initiative level is baked directly into the Worker's system instructions. It influences each of its "do I act or do I ask?" decisions.

3. Rules (guardrails)

In addition to initiative, rules set precise limits per action type: always, ask or never. They're detailed in the dedicated Rules & guardrails article — just remember here that they live alongside skills in the Behavior Studio.

4. "Discover": how the Worker activates a tool

This is the most confusing mechanism on first contact, so let's clarify it.

The Worker does not load all the tools of all your integrations permanently — that would be heavy and confusing. Instead:

1. The Worker sees the list of connected integrations (e.g. Slack is connected).
2. When a task requires it, it "discovers" the corresponding skill — internally, it calls a tool such as discover_slack_skill.
3. This discover checks that the connection exists:
   • if it's missing → message "❌ Integration not connected. Connect it in Settings > Integrations."
   • if it's there → the skill activates and the real tools (slack_post_message, etc.) become available.
4. From there, the Worker can actually act on the tool.

What you see: in the activity log, a line appears such as "Slack activated" when the Worker draws the tool. This is normal and a good sign: it means it found and activated the right integration.

⚠️ Common blocker: if the Worker replies "this integration is not connected" when you thought you had connected it, it's the discover failing. Check in the Integrations tab that the connection is indeed active (toggle on) — see the Integrations article.

5. Skill vs integration vs tool vs rule — the right mental model

Remember: you configure skills (knowledge) and integrations (access); the Worker takes care, on its own, of discovering and calling the right tools at the right time, within its rules.