Agents

An Agent is an AI assistant you configure to understand your data. It knows which tables it can query, what the data means, how to interpret business terminology, and what rules to enforce. Once set up, it powers Ask Stella so your team can get accurate answers in plain English.

---

​

Browsing Agents

Navigate to Agents in the sidebar to see all agents in your workspace.

​

Display modes

Toggle between two views using the view switcher in the top-right:

---

​

Creating an Agent

Click New Agent from the Agents page. A modal collects the minimum required to get started: When selecting data sources, inactive connections are shown greyed out with an (Inactive) label. Each entry shows {source name} • {database}. Once created, you are taken directly to the agent’s setup page to configure everything else.

---

​

Setup — Information & Data Access

The Setup tab is the primary configuration page for an agent.

​

Information

​

Data source access

Controls exactly which data the agent can see.

• Multi-select dropdown — add or remove data source connections.
• Table-level scope — expand any connected source to choose specific tables. All tables are included by default.
• A count badge shows {X} sources · {Y} tables at a glance.

​

Handling missing schema

If a table’s schema hasn’t been generated, a warning appears. For each affected table:

A missing schema means the agent doesn’t know what the columns in that table mean. Resolve all schema warnings before expecting accurate answers.

---

​

Setup — Overview (Instructions)

The Overview is the agent’s core instruction set — a freeform text block that tells it:

• What this data represents
• The business domain it operates in
• How to interpret ambiguous questions
• Any context about data quirks or conventions

​

Quick templates

Supaboard offers domain-specific templates you can append to your overview. Templates are loaded based on the data sources you’ve connected. Click a template to add its suggested language.

The overview is the single most impactful thing you can write for an agent. A detailed, precise overview produces dramatically better answers than a vague one.

---

​

Setup — Access & Controls

​

Enable Data Access

Toggle — When on, Stella can drill into the actual data behind charts to provide deeper analysis and follow-up insights. When off, responses are limited to aggregated results.

​

Enable Self-Improvement

Toggle — When on, Stella can update table descriptions, create new metrics and dimensions based on conversation patterns, and refine its understanding over time. Changes are tracked in changesets so you can review and approve them before they take effect.

​

User Access

Controls which workspace members can chat with this agent.

• All users (default) — every workspace member with Stella access can use the agent.
• Specific users — only selected users can access the agent. You (the creator) and the workspace Owner are always included and cannot be removed.

Select users from the dropdown, which shows all workspace members with their roles. Your own entry is marked (you).

---

​

Knowledge Base

Navigate to the Knowledge tab to manage the agent’s document library. The knowledge base gives the agent context beyond what’s in the database — data dictionaries, business definitions, policy documents, or any reference material that helps it answer questions accurately.

​

Uploading files

Drag files onto the upload zone or click to browse. Supported formats: Multiple files can be selected at once.

​

Uploaded files table

Click a file name to preview its contents. Click the delete icon to remove it — the change is reflected immediately.

---

​

Master Ruleset

Navigate to the Master Ruleset tab to define hard rules that govern how the agent interprets and answers every query — these are not suggestions. Use rules to:

• Enforce terminology — “always call this field ‘Net Revenue’, not ‘Revenue’”
• Define business logic — “always filter to the current fiscal year unless the user specifies otherwise”
• Set formatting standards — “always express percentages to 2 decimal places”
• Prevent certain response types

​

Managing rules

Create a rule manually:

1. Click New Rule.
2. Enter a Title (short label) and Description (the rule in plain language).
3. Save.

Import rules from a file:

1. Click Browse File.
2. Upload a YAML or JSON file containing rule definitions.
3. Supaboard extracts rules from the file automatically.

Edit or delete a rule: Click the action menu on any rule row to open the editor. Changes can be undone within the editor before saving. Search rules by title or description using the search bar at the top of the table.

---

​

Table Configuration

Navigate to the Table Configuration tab to teach the agent what each table means — context that a SQL schema alone cannot convey.

​

Navigating tables

Use the two dropdowns at the top to select a data source, then a table within it. A warning icon appears next to sources with missing schema.

​

Per-table actions

​

Configuration fields

For each table, fill in four knowledge fields:

Table Configuration is the second most impactful investment after the Overview. Granularity and Gotchas in particular prevent the most common categories of incorrect answers.

---

​

Fine-Tuning (Verified Queries)

Navigate to the Fine Tuning tab to manage Verified Queries — pre-written, validated SQL that the agent reuses for known questions. When a user asks something that closely matches a verified query, Stella uses it directly rather than generating a new one. This guarantees:

• Correctness for high-stakes metrics
• Consistent answers across all users
• Faster responses (no generation step needed)

​

Adding a verified query

Click New Evaluation and fill in:

​

Managing verified queries

• Search by title or description.
• Click View on the query column to see the full SQL.
• Use the action menu to edit or delete a query.

---

​

Integrations

The Integrations section within Setup connects the agent to external platforms.

​

Slack (Business)

Connect the agent to a Slack workspace so team members can query it directly from Slack without opening Supaboard. Click Connect to start the OAuth flow. The connection is scoped to this specific agent.

​

Microsoft Teams (Business)

Coming soon.

---

​

Managing Agents

​

Duplicate

Open the action menu (three-dot icon) on any agent and select Duplicate. You’ll be prompted for a new name. The duplicate copies all configuration: data scope, overview, knowledge, ruleset, table config, and verified queries.

​

Delete

Open the action menu and select Delete. A confirmation is required. Deletion is permanent — existing conversations that used this agent are preserved but can no longer generate new responses from it.

​

Unsaved changes

Navigating away with unsaved edits triggers a confirmation prompt. Draft changes are encoded in the URL so you can copy the link and return to continue editing later.

---

​

Permissions

Agent-level access is separate from workspace role — even an Admin can be excluded from a specific agent if it has a restricted User Access list (see Setup — Access & Controls).

---

​

Agent Setup Checklist

Use this to get a new agent to production quality:

• Connect the right data sources and scope down to relevant tables only
• Write a detailed Overview — explain the business domain, data conventions, and any ambiguities
• Fill in Table Configuration for every table, especially Granularity and Gotchas
• Upload a Knowledge Base with data dictionaries and reference documents
• Define Master Ruleset rules for business-critical terminology and logic
• Add Verified Queries for your 5–10 most frequently asked questions
• Set User Access to the right people
• Enable Self-Improvement once you are satisfied with baseline answer quality