cursor-agent-orchestrator

Orchestrate concurrent cursor-agent processes with timeouts and JSONL logging.

3|1|Updated Dec 23, 2025
One-click install
npx skills add https://github.com/marcus/marcus-skills --skill cursor-agent-orchestrator
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cursor-agent-orchestrator
Source: https://github.com/marcus/marcus-skills/tree/main/skills/cursor-agent-orchestrator
Command: npx skills add https://github.com/marcus/marcus-skills --skill cursor-agent-orchestrator

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This orchestrator enables you to run multiple cursor-agent processes in parallel, capturing outputs, enforcing per-task timeouts, and logging structured results for analysis and audit trails.

Core Features & Use Cases

  • Parallel cursor-agent orchestration with configurable concurrency
  • Timeout enforcement and robust error handling
  • Structured logging (JSONL) and transcript capture
  • Graceful shutdown with optional heartbeat monitoring

Quick Start

Configure your workspace and run the CLI to start concurrent orchestrations.

Frequently Asked Questions about cursor-agent-orchestrator

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run multiple cursor-agent processes in parallel?

You can orchestrate parallel cursor-agent runs by automating subprocess management to process large task batches concurrently. The system coordinates concurrent agents across files, databases, or API work while enforcing per-task timeouts and aggregating results.

How do I capture and log structured output from concurrent subprocesses?

Capturing structured output from concurrent subprocesses is handled through JSONL logging and transcript capture. This records aggregated results and provides structured logs for analysis and audit trails during parallel orchestration.

Can I enforce timeouts on individual tasks during parallel orchestration?

Timeout enforcement for individual tasks is supported during parallel orchestration. The system coordinates concurrent agents and enforces per-task timeouts to ensure robust failure handling across batch processing workflows.

What is the best way to handle graceful shutdown for async subprocess management?

Graceful shutdown for async subprocess management is implemented with optional heartbeat monitoring. This ensures robust failure handling and cleanly terminates cursor-agent processes when stopping parallel task orchestration.

Does this parallel orchestration approach work for batch tasks across databases and APIs?

Batch tasks across files, databases, or API work are fully supported by the parallel orchestration design. It coordinates concurrent agents to process these large task batches reliably with robust error handling.

Why do my concurrent agent runs fail without structured logging?

Concurrent agent runs without structured logging fail due to lack of result aggregation and audit trails. Implementing JSONL logging and transcript capture provides the necessary failure handling and analysis data for reliable orchestration.