Are you an LLM? Read llms.txt for a summary of the docs, or llms-full.txt for the full context.
Skip to content

Concepts

This section explains the core building blocks of Orloj and how they fit together. Each concept page covers what a resource is, why it exists, how to configure it, and how it interacts with the rest of the system.

If you are new to Orloj, start with the Architecture Overview to understand the system's layers, then work through the concepts below.

At a Glance

                  TaskSchedule ──creates──▶ Task ◀──creates── TaskWebhook
                                             │
                                          triggers
                                             ▼
                                        AgentSystem
                                        ╱          ╲
                                   composes      composes
                                     ╱                ╲
                                Agent A ─────────── Agent B
                               ╱   │   ╲           ╱   │
                          calls  invokes reads  calls invokes
                            ╱      │    ╲       ╱      │
                   ModelEndpoint  Tool  Memory  │      │
                        │          │            │      │
                   resolves    resolves          │      │
                    auth via    auth via         │      │
                        ╲       ╱               │      │
                         Secret                 │      │
                                                │      │
              ┄┄┄┄┄┄┄┄ Governance ┄┄┄┄┄┄┄┄┄┄┄┄┄┤┄┄┄┄┄┄┤
              ┆                                 ┆      ┆
        AgentPolicy ┄┄ constrains ┄┄▶ Agent A, Agent B
        AgentRole   ┄┄ grants permissions to ┄▶ Agents
        ToolPermission ┄ controls access to ┄▶ Tools
 
              Worker ──claims and executes──▶ Task

Core Resources

Agents and Agent Systems -- Agents are declarative units of work backed by language models. Agent Systems compose agents into directed graphs (pipelines, hierarchies, swarm loops) that Orloj executes as coordinated workflows.

Tasks and Scheduling -- Tasks are requests to execute an Agent System. They carry input, track execution state through a well-defined lifecycle, and support cron-based scheduling and webhook-triggered creation.

Tools and Isolation -- Tools are external capabilities that agents invoke during execution. Orloj provides a standardized tool contract, four isolation backends (none, sandboxed, container, WASM), and configurable timeout and retry.

Model Routing -- ModelEndpoints decouple agents from specific model providers. Configure connections to OpenAI, Anthropic, Azure OpenAI, or Ollama, and bind agents to endpoints by reference.

Memory -- Memory gives agents the ability to store, retrieve, and search information using built-in tools. Operates in three layers: conversation history, task-scoped shared state, and persistent backends.

Governance and Policies -- AgentPolicy, AgentRole, and ToolPermission resources enforce authorization at the execution layer. The governance model is fail-closed: unauthorized tool calls are denied, not silently ignored.

Architecture and Execution

Architecture Overview -- The three-layer architecture: server, workers, and governance.

Execution and Messaging -- Graph routing, fan-out/fan-in, message lifecycle, ownership guarantees, and tool selection.

Starter Blueprints -- Ready-to-run pipeline, hierarchical, and swarm-loop topologies with example manifests.