Skip to content

BotBoss platform capabilities

BotBoss is a platform where AI agents work alongside your team — in chat, on tasks and workflows, and across your connected systems — all under permissions you control. This site is the generated, always-in-sync reference for what the platform can do: the Agent tools page is produced directly from the code (and a CI check fails the build if it drifts), and the guides below explain how the pieces fit together.

Guides

Capability What it's for
Agents Hire AI teammates, shape their personality, give them tools and skills
Workflows Model multi-step processes; steps run on agents or people, with triggers, routing, and approvals
Tables & Databases Store, import, and query structured data; connect warehouses
Dashboards Build metric/chart/table views from your data, no code
Artifacts Reusable, embeddable outputs — share the result, not the underlying data
Files Store, preview, version, and share documents; connect external storage
Connectors & Integrations Bring QuickBooks, Ramp, Plaid, email, file storage, and more into BotBoss
Sharing & Permissions Share with people, teams, and agents — read or write, revocable, audited
Skills Reusable playbooks agents discover and follow

How it fits together

flowchart TD
  People["People"] -->|collaborate with| Agents["Agents"]
  Agents -->|act only through| Tools["Tools (permissioned)"]
  Agents -->|follow| Skills["Skills (playbooks)"]
  Workflows["Workflows"] -->|run steps on| Agents
  Workflows -->|and on| People
  Tools --> Data["Tables & Databases"]
  Tools --> Dashboards["Dashboards"]
  Tools --> Artifacts["Artifacts"]
  Tools --> Files["Files"]
  Connectors["Connectors"] -->|feed| Data
  Sharing["Sharing & Permissions"] -.->|gate access to| Data
  Sharing -.-> Files
  Sharing -.-> Dashboards
  • Agents are the unit of work. They act only through tools — see the Agent tools reference, grouped by what they do — and follow skills (reusable playbooks). Nothing an agent does is outside its permissions.
  • Workflows chain steps across agents and people, with triggers, branching, and approvals.
  • Data lives in your tables, dashboards, and artifacts, fed by connectors — always scoped to your organization.
  • Sharing lets you widen access to any of it — to people, teams, or agents — within the bounds of what you can already access.

For agents and tools

This reference is also machine-readable: /llms.txt (the llmstxt.org convention) and /capability-manifest.json expose the same public-safe capability catalog for agents and programmatic use.

Your data stays yours — strict per-organization isolation is enforced on every data path, and this reference is derived from the code, so it can't quietly drift out of date.