Skip to content

Agents and automation

This chapter documents the agent/task automation layer: custom agents, background agents, task tool constants, subagent lifecycle hooks, slash-command automation, ultrareview, and auto-mode.

Read this chapter when the question is: how does Claude Code delegate work, run subagents, or automate runtime behavior?

Source-anchor policy

This page is a chapter guide. Linked implementation pages carry concrete cli.js anchors.

Semantic alias	Minified anchor	Scope
Agents/automation chapter	N/A — navigation page	Groups custom agents, background agents, task tools, subagent hooks, slash commands, and automation commands.
Agent implementation pages	See linked source-anchor tables	Concrete bundle anchors live in destination pages.

Automation map

flowchart TD
    Root[Root command] --> AgentsCmd[agents command]
    Root --> AgentsFlag[--agents JSON]
    Runtime[Session runtime] --> TaskTools[TaskCreate / TaskGet / TaskList / TaskUpdate]
    Runtime --> Slash[slash commands]
    Runtime --> Hooks[Subagent and task hooks]
    AgentsFlag --> Custom[Custom agents]
    AgentsCmd --> Background[Background agents]
    TaskTools --> Subagents[Subagents / tasks]
    Slash --> Automation[command automation]

Primary reading order

Order	Page	Automation question answered
1	Agents, tasks, and subagents	Which command/flag/tool/hook surfaces define custom agents, tasks, background agents, and hosted review, and how do `TaskCreate`/`TaskGet`/`TaskList`/`TaskUpdate`, subagent hooks, cron scheduling, and `ultrareview` preflight work?
2	Agent runtime, scheduling, and completion	How are agent families designed, how are tasks scheduled, how is completion detected, and how do timed/cron tasks work?
3	Slash commands and automation	Which slash-command, hook, skill, and auto-mode surfaces automate behavior around the main session?
4	Agent and automation architecture	How are custom agents, tasks, slash commands, `auto-mode`, and hosted review orchestrated over the same runtime without parallel loops?

Handoffs

Custom-agent prompt/context inputs are documented in Context and model loop.
Tool permissions and hooks are documented in Tools, integrations, and security.
Remote/hosted session state is documented in Sessions, persistence, and remote.
Task/agent communication protocol families are summarized in Runtime communication protocols.

Created and maintained by Yingting Huang.