Orchestration

Launch, monitor and resume Claude Code or Codex runs through the Agent Monitor API

Review past dashboard runs and link them to their saved Claude Code sessions

Find bottlenecks in agent workflows, tool flow and subagent delegation patterns

Build a report on agent orchestration, delegation, tool flow and concurrency

Implement tickets with bounded worker subagents, then integrate their results

Spin up coordinated agents with one API call to scope, build and ship software

Install the CORAL CLI and run your first experiment on existing code

Author a CORAL task with task.yaml, seed code and a grader for automated scoring

Launch, monitor and steer CORAL agent experiments from the command line and dashboard

Run iterative evidence reviews between a research lead agent and an auditor agent

Run 20 persona agents in parallel on a goal and merge their views into one decision

Run a council of CLI planners that draft, anonymize and judge one merged plan

Coordinate parallel work through the explicit /parallel command

Run a set of tasks in parallel through the explicit /parallel-task command

Run a parallel task variant on explicit /parallel-task-spark requests

Launch a large agent swarm only on an explicit /super-swarm-spark command

Plan dependency-aware work splits for parallel multi-agent implementation

Run JSON-defined command stages in sequence or parallel with timeouts and failure rules

Route requests to Claude Code, Codex, Gemini CLI and other ACP coding agents