Skip to main content
Use the composer’s + menu for attachments, reusable instructions, connections, knowledge, and write controls. Choose the AI model separately from the model selector beside the composer.
Composer menu with Attach file, Add to message, Connections, Write actions, and Automatic knowledge

The current composer menu. Sample workspace shown.

​
The composer + menu

Other controls, including web search and Agent progress, depend on the conversation and selected model’s capabilities.

​
Connections (tools)

Connections are the external tools the assistant may use to do real work — querying NetSuite, calling other connectors, and Erstan’s built-in tools. Open Connections from the + menu to choose how tools are selected.
In Auto mode, Erstan discovers and picks the right tools for each message on its own — the panel describes this as “Connection Hub discovers tools per message.” This is the default and the best choice for everyday chat: you ask a question, and Erstan reaches for the NetSuite or connector tools it needs to answer it.
If a tool the assistant needs isn’t connected yet, you’ll be prompted to set it up — owners and admins connect NetSuite and other systems from Connectors. NetSuite uses per-user authorization, so if your own NetSuite user and role aren’t authorized yet, an Authorization Required card appears in the transcript and you can authorize without leaving chat; see Authorize your NetSuite access.
Connecting a tool that can change NetSuite data does not bypass safety. Write-capable actions follow the configured tool policy and conversation write mode. A denied tool remains blocked. Review write safety before enabling broad tool access.

​
Knowledge

Knowledge grounds answers in content selected by enabled sources. Open Knowledge from your user menu, click Add source, and choose the content an Agent should search. Use Continue to review it before adding the source, then attach the named source to the Agent. Team scope controls narrow retrieval within those selections; indexing alone never makes content available.
Select files, documents, or bounded folders, such as a vendor policy or a chart of accounts mapping. Results require both source access and access to the original content. See Agent knowledge.
Workspace owners can add a source with specific Erstan-managed NetSuite references. Clear release, version, feature, or product-update questions can use those selected references. There is no automatic subscription, and future catalog additions remain unselected.
In the source menu, use Pause sync to stop updates while keeping current content searchable when search is enabled. Disable search stops searches through the source independently of sync. Remove source removes the selection while keeping original content. Another enabled source can still include the same material.
Knowledge is for context the assistant should read. To have it act on NetSuite — pulling live invoices, balances, or records — that comes through Connections, not Knowledge.
The Web search toggle lets the model look up information on the public web during a reply, which is useful for general context that isn’t in your NetSuite account or your team files. It appears in the + menu only when the selected provider and model support web search; if you don’t see it, choose an eligible model from the model selector and check the workspace tool settings.
Web search reaches outside your workspace. It does not read your NetSuite data — that always flows through Connections and Knowledge — and it has no bearing on approvals for write actions.

​
Image generation

With an OpenAI model selected, Image Generation appears under the OpenAI connection alongside Web Search and Code Interpreter. It is enabled by default and managed in the workspace’s OpenAI connector settings. Ask Erstan to generate an image; the result appears as an image attachment you can preview or download. Availability also depends on the selected model’s tool policy and OpenAI account access.

​
Model settings

Open the model selector beside the composer. Search the available models or group them By provider or By purpose. Choose the workspace default, a specific model, or Auto when offered.
Model picker with search, provider grouping, workspace default, and reasoning effort

Model selection and reasoning effort. Your catalog may differ.

The picker shows only choices available to your workspace. For a model that supports it, use Effort to set the reasoning level. Higher effort can take longer. Browse models opens the wider model catalog. Chat can remember your selection. Workspace owners and admins manage AI access in Connectors → AI and defaults in Settings → AI. Managed access does not require your own provider key; customer keys are supported only where permitted. See AI providers and models.

​
Next steps

Chatting with Erstan

The basics of starting a chat and reading replies.

Approving AI actions

Review write actions before they run, and when to auto-approve.

Connect an LLM provider

Review managed access, customer keys, Auto, and available models.

Agent knowledge

Make team files searchable so chat and agents can use them.