Studio Assistant

The Studio Assistant is an AI helper available from any page in Vertesia Studio. Think of it as a collaborator who knows the product — it can design interactions, draft prompts, build content types, create agents, author processes, set up databases, and help administrators inspect and configure access control.

It lives behind the bot button in the bottom-right corner. Click it, a side panel opens, and you're in a chat. The assistant carries context about the page you're on, so asking "help me structure this prompt" on a prompt-editor page does the right thing without you having to explain which prompt.

Where it lives

• Present on every Studio page - toggle the Studio Assistant panel open/closed using the icon in the bottom right corner.
• Per-user and per-project persistence — the selected environment, model, and active conversation are restored when you return to the project in the same browser.

How it works

Three things make Studio Assistant different from a generic chat assistant:

1. Skill-driven tools. The assistant loads the current product guidance for an area before it uses that area's tools.
2. Page context. It knows which Studio page and resource you are viewing, so you can ask about “this interaction” or “this collection” without copying an ID.
3. Configurable environment and model. A small configuration strip lets you choose which execution environment and model drive the assistant. It defaults to the project's default environment.

What it's good at

• Authoring. "Design an interaction that extracts line items from invoices", "draft a prompt for classifying support tickets", "create a content type for legal contracts". The assistant loads the applicable product guidance before proposing a schema or changing a resource.
• Explaining. "What does this prompt do", "why is this agent failing to call the tool I declared", "what are my options for routing in a process". Each domain skill carries the current best-practice guidance.
• Orchestrating authoring. "Build me a contract-review process" prompts the assistant to draft, validate, and save a process once it has enough context.

Access-control administration

For administrators, Studio Assistant can:

• list the current system RBAC and content ABAC roles;
• inspect project access-control entries;
• create, update, or remove fixed role grants and Principal Set/Resource Set rules after confirmation;
• dry-run a proposed document or collection rule against a principal context and sample objects.

The dry run validates one rule's logic. It does not calculate other grants, direct object security, project defaults, or administrator bypasses. After saving a policy, always retest with a freshly issued token. See Access control overview and Dynamic access rules.

Process authoring

For process work, the assistant can:

• draft a full process definition from your description;
• validate a process draft before saving;
• create or update a draft process definition;
• publish only after explicit user confirmation;
• explain a selected process, node, or guard;
• improve the process-board layout without changing runtime behavior.

When improving layout from a screenshot, the image is only visual evidence. The assistant still needs the actual process definition and should change only titles, descriptions, labels, grouping, and visual order—not nodes, transitions, guards, tools, or other runtime behavior.

When designing node behavior, ask the assistant to reuse the project's existing interactions, tools, and agents where possible. Create a custom interaction only when the node needs a distinct contract.

What it's not

• Not a replacement for the Agent Runner. For running a production agent against a known interaction, use the Agent Runner — you get the full conversation page, artifacts, observability, and restart/fork controls.
• Not for long-running batch work. The side panel is designed for interactive assistance. For a batch job or a scheduled process, author the process and run it.
• Not silent. It always uses the model you configured; there is no hidden background model selection. If you close the panel mid-turn, the active run continues and can be resumed later.

Invoking it

Click the bot button in the bottom-right corner, then type. The assistant sees the page context automatically, so you can refer to "this interaction" or "this collection" without naming an ID.

Starting a new chat

The panel has a New chat control that starts a fresh conversation. Save any important output or run reference before starting over.

Steering it

A few patterns make the assistant noticeably more useful:

• Name the area and desired outcome. "Design the interaction contract for this extractor" gives the assistant a clear starting point.
• Give it enough context upfront. If you want an interaction designed for a specific use case, state the inputs and expected outputs in the first message.
• Ask for tool calls explicitly. "Create the interaction" is clearer than "could you make the interaction". The assistant is happy to narrate; if you want action, ask for action.
• Correct it out loud. When it drifts, say so. A direct correction is the fastest way to steer it.

See also

• Agents — the full-page agent runner for production agents.
• Authoring Processes — process authoring with the Studio Assistant and the learn_process_design skill.
• Interactions — what the assistant helps you author.