Skip to main content

Agent Platform

Module: operations:agents | Role: Any member (create), platform admin only (promote/demote) | Licence: Pro+ Tools: create_agent, update_agent, remove_agent, list_agents, view_agent, list_agent_tool_groups

The agent platform enables creation and deployment of specialized AI agents. Eve is the single orchestrator — users interact with Eve, who dispatches to specialized agents as needed. Agents never call each other directly.

Key Concepts​

ConceptDescription
AgentA specialized AI with a prompt, tools, and optional triggers
Agent TeamOrganizational grouping of agents under a module with a shared mandate
ModuleCommercial feature gate that controls which tools and agents are available
DispatchEve calls an agent via a dispatch tool (e.g., ask_policy_researcher)
LANESDesign vocabulary for agent roles (see below)

LANES Patterns​

Every agent follows one of six patterns that determine its role and default behavior:

PatternRoleExamples
LibrarianStore, index, retrieve knowledgePolicy Librarian (vault operations)
AnalystRead sources, create briefings/reportsPolicy Analyst (digests, summaries)
NavigatorTrack work, flag risks, generate statusPolicy Navigator (watchlist, deadlines)
ExplorerMonitor external signals, surface what mattersPolicy Explorer (regulatory scanning)
StewardWatch internal systems, detect deviationsAgent Steward (health monitoring)
BespokeUnique function not fitting other patternsCustom agents

LANES is not a class hierarchy — it provides prompt templates and tool defaults.

Agent Origins​

OriginCreated ByModifiableVisible To
BuiltinCode (engineering)NoAll workspaces with module
WorkspaceWorkspace membersYes (by creator)Current workspace only
PromotedAdmin promotionAdmin onlyAll workspaces with module

Agent Modes​

ModeBehavior
ExecuteTakes action (creates scenarios, writes to vault, generates reports)
TrackMonitors and reports (watchlist updates, status digests, alerts)

Scoping & Access Control​

The platform uses a five-layer access model (enforced by check_agent_access()):

  1. Module gate — Is the required module enabled for this workspace?
  2. Workspace gate — Does the agent belong to this workspace (or is it builtin/promoted)?
  3. Origin gate — Is the agent's origin compatible with the operation?
  4. User visibility gate — For restricted agents, is the user in visible_to?
  5. Conversation scope gate — OAuth and signed-in sessions see private agents; shared connections see only shared agents

Visibility​

  • workspace (default): All workspace members can see and interact
  • restricted: Only users listed in visible_to can see and interact

Triggers​

Agents can run autonomously on schedule or in response to events:

TypeDescriptionExample
CronScheduled execution"0 6 * * *" (daily at 6 AM)
EventTriggered by system eventsModule enabled, data refresh
WebhookExternal HTTP triggerPartner API callback
ManualUser-initiated via Eve"Run the regulatory scan"

Delivery Channels​

Triggered agents can push their output directly to the user who created the trigger:

ChannelDescription
slack_dmDirect message via the workspace's Slack bot
emailEmail to the user's registered address

Configure per-trigger: "delivery": ["slack_dm", "email"]. See Creating Agents for details.