span-of-control

What problem does it solve?

Claude can take on complex or parallel tasks without visible limits, often leading to degraded output or unnoticed quality risks. This skill makes Claude declare its reliable operating span before starting work, so you can choose how to proceed.

Core Features & Use Cases

• Visible span assessment across the three dimensions: parallel task count, task complexity, and coordination load.
• Provides clear options when the span is exceeded: sequential execution, scoped parallelism, or full parallel with quality risk notes.
• Integrates with structuring skills (e.g., contract-first-decomposition) to extend span through better scoping and verification.
• Suitable for long-running agentic workflows and multi-tool orchestration in complex sessions.

Quick Start

Install the SKILL.md into your Claude skills directory, then Claude will assess span before starting complex tasks.