Configure an agent
Edit an agent's instructions, model, tools, memory, and connected accounts. Everything lives in the agent details panel.
Every agent in your workspace has a details panel where you can edit how it behaves, what it can do, and what it remembers. Open it by clicking the agent's name in any conversation header.
Visibility (private vs workspace) controls who can use an agent — and a shared agent acts through your connected apps. See Agent scoping & access.
Opening the details panel
Name and description
The name helps Nebula decide when to delegate work to this agent — keep it descriptive. The short description appears in mentions, and Nebula uses it when routing tasks.
Visibility
Each agent in a workspace is either:
| Workspace | Private | |
|---|---|---|
| Who can see it | Every member of the workspace | Just you |
| Who can use it | Every member | Just you |
| Best for | Team utilities, shared agents | Personal experiments inside a team workspace |
Model
Click the model badge (e.g. "Claude Sonnet") next to the agent's name to open the model picker. Pick one and it applies immediately.
Choose the model based on what the agent does — heavy code review (Claude Sonnet), big-context research (Gemini Pro), bulk classification (Gemini Flash).
Instructions (system prompt)
Instructions define the agent's role, voice, and rules. They're sent with every message.
Click Edit next to Instructions in the details panel, or ask Nebula to update them in chat.
Update my support triage agent to always classify issues as P1–P4 and tag the relevant team.
Done — updated the instructions to include priority classification (P1–P4) and team tagging for all incoming issues.
Memory
Agents remember useful facts across conversations — preferences, resource mappings, project context — so you don't repeat yourself. Memory is separate from any single conversation's history.
After an agent completes a task with tool calls, Nebula extracts useful facts in the background. You don't have to do anything.
Just ask:
Remember that staging is at staging.acme.com.
Saved — I'll remember your staging URL.
Pin a memory to one channel so it doesn't leak elsewhere:
Just for this channel: the client is Acme Corp.
Noted — scoped to this channel only.
Forgetting or correcting works the same way — ask the agent to update or forget it. Memories also live in the Knowledge tab, where you can browse, edit, or clear them.
Tools
Tools are the things the agent can actually do — web search, code, browsing, app actions. They're grouped by toolkit in the panel; expand any toolkit to see what it includes.
See Agent tools for the full catalog and what each tool does.
Accounts and variables
Third-party connections this agent uses (Slack, GitHub, Gmail). A green check means active. When you've connected the same app twice in the workspace (two Slack workspaces, say), a dropdown lets you pick which one this agent should use.
Secrets and config values some tools require — API keys, tokens. Status badges show Set (ready) or Missing (action needed). Always set them through the form in the panel; never paste a secret into a chat message.
Triggers
Any automations attached to this agent appear here with their schedule or event type. Click one to edit, hover and click × to remove.
Deleting an agent
Scroll to the bottom of the panel and click Delete agent. The agent's configuration is removed; its past conversations stay in the workspace history.
Platform agents (marked with a Platform badge) and the main Nebula agent can't be deleted.
Related
Create Custom Agents
Build custom AI agents in Nebula to automate specific tasks, workflows, or app integrations. Create them from the sidebar or in chat.
What agents can do
Explore the built-in tools, connected app actions, and the permissions model for AI agents in your Nebula workspace. From web search to code.