Skip to main content

Platform

SOAT's functionality is organized into modules — named resources exposed through the REST API, the MCP server, the CLI, and the SDK. Each module page describes what the resource does, its data model, key concepts, and usage examples on every client surface.

The modules fall into seven groups:

Identity & Access​

Who can do what, and with which credentials.

  • Users — accounts, roles, and authentication
  • Projects — the primary resource boundary; almost everything belongs to a project
  • IAM & Policies — how permissions are evaluated
  • Policies — reusable policy documents granting resource:Action permissions
  • API Keys — project-scoped and personal keys with policy attachments
  • OAuth — the OAuth flow used by MCP connectors

Storage & Retrieval​

Project-scoped data and semantic search.

  • Files — binary file storage
  • Documents — structured text content with ingestion
  • Embeddings — pgvector embeddings and semantic search
  • Ingestion Rules — automatic processing of uploaded content
  • Knowledge — unified search across documents and memory entries
  • Memories — durable context stores for agents

Agents & Conversations​

The generation engine and its building blocks.

  • AI Providers — LLM provider connections and models
  • Agents — configurable agents with tools and multi-step reasoning
  • Tools — HTTP, MCP, client-side, and SOAT-platform tools
  • Sessions — the 1↔1 user/agent interface
  • Conversations — the multi-party message engine
  • Chats — raw LLM completions without an agent
  • Actors — participant identities in conversations
  • Generations — generation records and async jobs
  • Chains — continuation chains: the linked tree a resumed turn grows into, and the ceilings that stop it

Orchestration & Automation​

Composing agents into workflows and reacting to events.

  • Orchestrations — deterministic multi-agent graphs with typed state (a pipeline that ends)
  • Workflows & Tasks — state-machine definitions and the durable, stateful work items that live in them (an entity that moves between states, including backward)
  • Triggers — scheduled and on-demand flow execution
  • Webhooks — HMAC-signed event delivery to external systems
  • Approvals — human-decision queue for agent-proposed actions
  • Guardrails — action-class policies that classify each agent tool call

Declarative Deployment​

  • Formations — define full agent stacks (providers, memories, tools, agents) in JSON/YAML and deploy them with dependency-aware provisioning. See the Formation Types reference for every resource type.

Improvement & Adaptation​

How the system changes, and what proves a change was an improvement. See The Layers of an Agent System for how this group fits the rest of the platform.

  • Evaluations — datasets, scorers, and scored runs comparable against a baseline (coming soon)
  • Agent versions — append-only config history with staged canary rollout
  • Approvals recurrence view — what human correction keeps coming back, and what to do with it

Operations​

Observability and runtime configuration.

  • Traces — per-generation trace records: tool calls, latency, token usage
  • Usage — usage metering and pricing
  • Secrets — encrypted secrets for provider keys and tool credentials
  • Docs — MCP-only tools that give agents access to SOAT documentation

See the Permissions Reference for the full list of IAM action strings across all modules.