pi-subagent

Orchestrate task delegation to isolated subprocesses for parallel or sequential execution.

114|12|Updated Feb 11, 2026
One-click install
npx skills add https://github.com/espennilsen/pi --skill pi-subagent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pi-subagent
Source: https://github.com/espennilsen/pi/tree/main/extensions/pi-subagent/skills/pi-subagent
Command: npx skills add https://github.com/espennilsen/pi --skill pi-subagent

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill streamlines complex workflows by allowing you to delegate tasks to specialized, isolated subagents, enabling parallel execution, sequential pipelines, and improved context management.

Core Features & Use Cases

  • Parallel Task Execution: Run multiple independent tasks concurrently across different agents.
  • Sequential Task Chaining: Create pipelines where the output of one agent feeds into the next.
  • Context Isolation: Ensure subagents operate with a clean context, preventing interference.
  • Use Case: Automate a code review process by having a 'scout' agent identify potential issues, a 'planner' agent prioritize them, and a 'worker' agent implement the fixes, all in a controlled sequence.

Quick Start

Use the pi-subagent skill to run a scout agent to find all REST endpoints in the src directory.

Frequently Asked Questions about pi-subagent

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

FAQPage Schema
How do I delegate tasks to parallel agents for concurrent execution?

You can delegate tasks to parallel agents by orchestrating task delegation to isolated subprocesses, which enables concurrent execution of multiple independent tasks across different specialized agents. This approach streamlines complex workflows by managing agent specialization and context isolation.

How does context isolation work when chaining sequential agent pipelines?

Context isolation in sequential agent pipelines ensures that each subagent operates with a clean context, preventing interference from previous tasks. It manages task chaining by feeding the output of one isolated subprocess directly into the next agent for controlled sequential execution.

Can I use per-task overrides to fine-tune model thinking level for specific agents?

Yes, you can use per-task overrides to fine-tune execution by specifying different model configurations, thinking levels, extensions, and skills for individual isolated subprocesses. This allows precise control over how each specialized subagent processes its delegated task.

What is the best way to automate a multi-step code review process with specialized agents?

The best way to automate a code review process is sequential task chaining using specialized agents: a scout agent identifies issues, a planner agent prioritizes them, and a worker agent implements fixes. This controlled pipeline ensures context isolation and prevents workflow interference.

Do I need any external dependencies to run isolated subprocesses for parallel workflows?

No external dependencies are required to run isolated subprocesses for parallel workflows. The skill manages agent specialization, context isolation, and task chaining internally, allowing you to execute complex pipelines without installing additional packages or external libraries.