The Brief¶

Preview

Framework v2 is not yet available to all accounts.

The Brief is the one document you write to describe an agent. It is what the console shows on the agent's Definition page, what you export and import, and what an Agent Template copies onto a new agent. In v2 it is also, word for word, the text the agent reads. There is no hidden text added to it at runtime.

The Brief has six sections. Four apply on every turn; Playbooks apply one at a time; Content applies when a Playbook or a Policy calls for it.

Section	What it holds	Applies
Identity	Who the agent is: name, company, contact details, branding.	Every turn
Personality	How the agent sounds.	Every turn
Guardrails	What the agent never does or says.	Every turn
Objectives	What a successful conversation achieves, in priority order, each with an outcome.	Every turn
Playbooks	What the agent does, one situation at a time.	The active one in full; the others by name and when-to-use
Content	Message templates: exact copy you have reviewed.	When referenced

Identity, Personality, Guardrails¶

Identity is who the agent is; the agent introduces itself from it and gives out contact details from it. Personality is the voice, applied everywhere; a Playbook can ask for a situational posture ("be brief while booking") without changing the voice. Guardrails are what the agent never does: topics to avoid, claims not to make, disclaimers to include. Guardrails steer the agent's own judgment; rules the platform must enforce regardless of judgment are Policies.

For Mia: Identity is Glow Aesthetics, with the clinic's phone number and website. Personality is warm and unhurried. Guardrails say never to give medical advice or quote a price for a treatment that is not in the catalog.

Objectives, with outcomes¶

An Objective is what success looks like at the end of a conversation, ranked so the agent knows which matters most when they pull against each other. In v2 each Objective can name the outcome that proves it was reached: an event the platform records, such as a call being booked. When it does, completion is counted from what actually happened, not judged from the transcript afterwards.

Mia has two Objectives: A consultation call is booked (outcome: a call booked through Scheduling) and The person's treatment interest is known (outcome: the treatment-interest field on their Person record is filled in).

Playbooks¶

A Playbook is what the agent does in one kind of situation. Every Playbook has four parts.

Part	What it is
Name	The label you see in the console and in the turn record.
When to use	One or two sentences describing the situation. This is how the Playbook is chosen.
Body	Instructions the agent reads while this Playbook is active.
Uses	The capabilities this Playbook relies on.

A Playbook that follows a fixed path (booking, consent, intake) also carries a Flow; see Playbooks and Flows.

Mia's booking Playbook, in the portable document format:

- name: Book a consultation call
  when: The person wants to book, schedule, or set up a call.
  uses: [scheduling, people]
  body: >-
    Confirm the person's name and email if we do not have them, offer
    times from the team's calendar, confirm before booking, send the
    booking confirmation, then return to general help.

How the active Playbook is chosen¶

At the start of every turn, Anychat reads each Playbook's when-to-use line against the conversation so far and picks the one that fits. The choice is made before the agent writes anything and is written into the turn record, so you can always see which Playbook was active and why. When no Playbook fits, the default Playbook (Mia's Welcome) is active. A Playbook whose capabilities are not ready is not a candidate; see Capabilities.

You do not write exit rules. When a later message fits a different Playbook, that one becomes active; when a task is done, the agent returns to the default. A Playbook's body can say how it wraps up ("send the confirmation, then return to general help").

What the agent reads from the Playbooks section¶

On any turn the agent reads the active Playbook's full body plus an index of every other Playbook: name and when-to-use only. It does not read the bodies of Playbooks that are not active. So the Definition page is the most the agent could ever see, and each turn is a predictable slice of it: the whole Brief, minus the inactive bodies. Preview what the agent sees shows exactly that slice for a chosen Playbook and channel.

While Mia is answering a treatment question, she reads the Answer treatment questions body in full and knows only that Book a consultation call exists and when it applies. That is enough for the handoff to happen and keeps every turn short and focused.

What `uses` means¶

uses names the capabilities a Playbook relies on. It does two things. It sets what the agent may do while the Playbook is active: Mia can offer times and book only in a Playbook that uses Scheduling. And it lets Anychat check the agent before a conversation happens: a Playbook that uses Scheduling is ready only when Scheduling is ready, and a Playbook whose body talks about booking but does not use Scheduling is flagged by lint. Capabilities not named in uses are not offered during that Playbook, which keeps the agent from doing something the situation did not call for.

Content¶

Message templates are exact copy you have reviewed, with placeholders filled from the Person record and the agent's identity ({person.firstName}, {agent.companyName}). Where a Playbook, Flow, or Policy references a template, the platform sends the template's text verbatim; the agent does not rewrite it. Mia's booking confirmation and her outreach opening are both templates.