generic-subagent

Delegate context-expensive tasks to read-only subagents and return concise digests.

Updated Mar 5, 2026
One-click install
npx skills add https://github.com/Nathanto-ai/liza-private --skill generic-subagent-nathanto-ai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: generic-subagent
Source: https://github.com/Nathanto-ai/liza-private/tree/main/skills/generic-subagent
Command: npx skills add https://github.com/Nathanto-ai/liza-private --skill generic-subagent-nathanto-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Subagents offload context-expensive work to callers, delivering concise digests instead of full traces to reduce context load and preserve focus.

Core Features & Use Cases

  • Default read-only subagents for research, analysis, and summarization to minimize side effects.
  • READ-WRITE opt-in to allow state changes within explicitly declared scopes.
  • Structured digest outputs that enable downstream tasks to proceed without exposing full histories.
  • Contract-driven delegation that improves reliability in complex multi-step workflows.

Quick Start

Provide a task objective and ask for a concise digest of the results.

Frequently Asked Questions about generic-subagent

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

FAQPage Schema
How do I delegate context-expensive research to subagents without losing focus?

Default subagents operate in a read-only mode to minimize side effects during research, analysis, and summarization. You must explicitly declare a READ-WRITE scope if your delegation requires state changes within the workflow.

What is the best way to manage context load when automating multi-step analysis?

Subagent delegation is suitable for bounded inputs like research, analysis, and summarization where outputs can be provided as digests. It is not suitable for open-ended tasks with unbounded inputs or unstructured outputs.

Can I use subagents to allow state changes in my automation workflow?

Yes, you can use subagents to allow state changes by opting into a READ-WRITE scope. This explicitly declares the permissions for the delegation, overriding the default read-only mode to enable controlled state modifications.

How does returning a digest instead of a full trace improve multi-step workflows?

Returning a digest instead of a full trace improves multi-step workflows by reducing context load and preserving focus. It enforces a contract that delivers structured results, enabling downstream tasks to proceed without exposing full execution histories.