Skip to main content

Overview

The Registry is the catalog of every AI agent that works inside your organization. It shows what each agent does, whether it’s currently running, and the capabilities it’s permitted to use. Open it from Agents → Registry.

How agents are grouped

Agents are organized into four groups by the kind of work they do:
GroupWhat these agents do
AssistantsConversational agents — the Chat Agent that answers questions, and the Automation Agent that helps you build and revise automations
ExtractionAgents that read documents and pull out structured data — Contract, Statement, Contract Repair, and Statement Repair agents
ReconciliationAgents that match and enrich your catalog — asset matching, catalog search, Spotify/Discogs/MLC/Chartmetric enrichment, and contract grouping, relationships, and consolidation
KnowledgeThe Brain Maintainer, which extracts and maintains your organization’s Brain

Reading the table

Each group lists its agents with three columns:
ColumnWhat it shows
AgentThe agent’s name and a short description of what it does
StatusWhether the agent is Enabled, Disabled, or Planned (coming soon)
ConfigurationA Configure button, a Managed badge, or Not available

Status

  • Enabled — the agent is active for your organization.
  • Disabled — the agent is available but currently turned off. For example, the Contract Repair Agent ships disabled by default. The Statement Repair Agent also appears here, but stays turned off for now — automated statement repair is coming soon.
  • Planned — the agent is on the roadmap but not yet listed for your organization.

Configuration

  • Configure — this agent can be adjusted. Click to open its settings panel.
  • Managed — this agent runs with fixed settings maintained by Royaltyport and can’t be changed.
  • Not available — shown for planned agents that aren’t running yet.

Configuring an agent

Click Configure on an agent to open a side panel where you can:
  • Enable or disable the agent for your organization (only for agents that can be turned off).
  • Adjust its capabilities per role. Some agents work in more than one role (for example, the Chat Agent has a Main role that plans answers and an Explore role that investigates files and records). Each role lists the capabilities it can use.

Agent capabilities

The capabilities you can grant control how an agent interacts with your organization’s knowledge:
CapabilityWhat it allows
Read BrainRetrieve organization facts, decisions, interpretations, policies, and context
Explore relationshipsTraverse the governed connections between organizational concepts
Read PlaybooksRetrieve the organization and project guidance relevant to the agent’s work
Open supporting evidenceFollow citations from Brain results back to their supporting sources
Propose Brain changesSubmit evidence-backed Brain observations to review — without publishing them
Propose Playbook changesSubmit Playbook improvements to review — without activating them
Agents can only ever propose changes to your Brain and Playbooks. Nothing an agent submits becomes active until a person reviews and approves it — see Brain and Playbooks.
Only organization admins can change agent settings. Per-organization agent configuration is rolling out — until it’s fully enabled, the Registry reflects Royaltyport’s default settings for each agent.

Brain

The organizational knowledge agents read from.

Playbooks

The procedural guidance that steers extraction agents.

AI Chat

The Chat Agent in action inside a project.

Sources

Feed documents and data into the Brain.