All use cases
Use case 03Team collaboration
Agents with a job each, in the chats your team already uses.
Split work across agents, keep automations versioned, and decide when the assistant speaks in a group.
01How it works
The mechanisms involved.
- An agent per job
- Each agent has its own personality, tool allowlist, memory scope, and model preferences. Its AGENTS.md says when to hand work to another agent.Where the docs say so: An agent per job
- Workflows from templates
- Start a workflow from a template in the Dashboard, or describe it and let the agent generate one. Every change is versioned, and an agent-authored workflow needs your approval.Where the docs say so: Workflows from templates
- Standup digest
- A heartbeat collects yesterday's completed tasks into a short summary.Where the docs say so: Standup digest
- Group chat controls
- Per-channel policies (mention-only, allowlist, all messages, disabled) decide when the assistant answers in a group, and Slack threads, Discord forum posts and Telegram topics each get an isolated session.Where the docs say so: Group chat controls
02Limits
What this doesn't cover.
- 01
Memory is scoped to a user or to one agent. User-scoped memory is shared across the agents you own, not across the people on your team.
Source for limit 1
Run it on your own infrastructure.
Docker Compose, a model key, a channel. The Personal tier is free for individual, non-commercial use.