Visibility

Description

The Visibility setting lets you control whether learners can open the exact source documents an agent retrieved to answer their questions. When visibility is on, the chat shows a Retrieved Documents side panel where users can click and read those files. When visibility is off, the agent still uses the resource for answers, but users cannot see or open it—ideal for proprietary, assessment, or advanced materials.

Target Audience

Administrator

Features

Per‑Resource Eye Icon

An eye icon (👁️ = visible, 👁️‍🗨️ = hidden) next to every dataset entry toggles learner access with a single click.

Side‑Panel Source List

Visible resources appear in a side panel after the agent answers, giving learners direct access to primary materials.

Non‑Destructive Control

Hiding a document never removes it from the agent’s training; it only affects learner visibility.

Instant Updates

Changes to the eye icon apply immediately—no retraining required unless you re‑enable a previously hidden file.

How to Use (step by step)

Open the Datasets Tab

Click the agent’s name → Settings → Datasets

Locate the Resource

Scroll or search for the file, URL, or cloud link you want to adjust

Toggle Visibility

👁️ (no slash) → learners see the document
👁️‍🗨️ (with slash) → learners cannot open the document

Click the icon to switch states:

Turning Off hides the resource (agent still uses it)
Turning On reveals it again in the side panel

Verify in Chat (Optional)

Ask a question related to the resource
If visibility is On, the source appears in Retrieved Documents
If Off, the panel shows no link to that file

Pedagogical Use Cases

Transparent Learning & Citation Practice

Make scholarly articles visible so users can read and cite original sources.

Scaffolded Disclosure

Start a unit with visibility off to prevent spoilers, then toggle on for deeper study.

Assessment Security

Hide answer keys or test banks while still letting the agent reference them for feedback.

Proprietary Material Control

Keep licensed or sensitive documents hidden for external cohorts.

Reveal supplementary readings only to learners who ask advanced questions, preserving focus for beginners.

By deciding which datasets are visible, you strike the perfect balance between learner transparency and content control—all while keeping the agent’s full knowledge base intact.