Granular Permissions for Conversation AI Agents in Sparkforge

By Sparkforge Support Team ·Last reviewed May 3, 2026

Granular permissions in Sparkforge let you decide which staff can manage Conversation AI Agents, edit goals, or train bots. Toggles live under each user's Roles and Permissions in My Staff.

Conversation AI Agents drive automated chat replies, qualification flows, and bot goals. Granular permissions in Sparkforge let you scope who can build, edit, train, or only view those agents on a per-user basis.

Where the permissions live

From your sub-account, click Settings in the lower left corner.
Open My Staff from the left navigation.
Find the staff member and click the pencil Edit icon.
Open Roles and Permissions from the secondary nav.
Scroll to the AI Agents section.
Toggle each permission on or off and click Save.

The four toggles

AI Agents (parent control)

The master switch. Disabling it hides the entire Conversation AI section from the user and turns off the dependent toggles below.

View and Manage Conversation AI Agents

Controls overall management: create, edit, and delete bots.

Enabled: full access to create, edit, and delete agents. Activates the Goals and Training toggles below.
Disabled: the Conversation AI menu is hidden and the user cannot create or edit bots. Dependent toggles are forced off.

View and Manage Conversation AI Agent Goals

Controls editing of bot goals — the instructions and response logic that shape behavior.

Enabled: the user can edit Bot Goals on existing bots. Best for power users tuning logic.
Disabled: the Bot Goals tab is hidden during edit. Form-based bots can still have goals defined during creation, and trial access is unaffected.

View Conversation AI Agent Goals

Read-only visibility for users who need to see goals without changing them.

Enabled: the user can read configured goals, instructions, and actions, plus prompt details inside the AI response info drawer.
Disabled: the Bot Goals tab is hidden entirely.

View and Manage Conversation AI Agent Training

Controls training data and example management for the agent.

Enabled: the user can upload, edit, or remove training examples and knowledge.
Disabled: training tabs are hidden and the user cannot push new training data.

Practical access patterns

Bot owner: all four toggles on.
Operator who handles handoffs: View Goals only. They can read bot logic without changing it.
Trainer or knowledge writer: View & Manage Training, but leave Manage Goals off.
External contractor: AI Agents parent off entirely until scope is defined.

Frequently Asked Questions

Where are the Conversation AI permissions configured?

Open Settings, go to My Staff, edit the user, open Roles and Permissions, and scroll to the AI Agents section.

What happens if I turn off the parent AI Agents toggle?

The user loses access to the Conversation AI menu entirely and the dependent toggles are disabled automatically.

Can a user view goals without editing them?

Yes. Enable View Conversation AI Agent Goals and leave View and Manage Goals off. The user gets read-only access.

Do form-based bots still let users define goals at creation?

Yes. Even with the Manage Goals toggle off, goals can still be defined during the form-based creation flow, and trial access remains.

Are training and goals separate permissions?

Yes. Goals control bot logic and instructions; training controls knowledge and examples. Each has its own toggle.