autosearch:router

Route autosearch tasks to 1–3 skill groups and 3–8 leaf skills.

40|6|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/0xmariowu/Autosearch --skill autosearch-router
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autosearch:router
Source: https://github.com/0xmariowu/Autosearch/tree/main/autosearch/skills/router
Command: npx skills add https://github.com/0xmariowu/Autosearch --skill autosearch-router

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The Runtime AI often loads many leaf SKILL.md files at session start. This Skill routes an autosearch task to the smallest relevant group of skills, loading only the matching leaf skills.

Core Features & Use Cases

  • Progressive-disclosure routing: identify 1–3 groups, then 3–8 leaf skills from those groups.
  • Lazily load: only the selected leaf SKILL.md when needed, reducing context size.
  • Robust fallback: if no groups match, degrade gracefully to generic web channels.

Quick Start

Provide a research query to see the router pick 1–3 groups and 3–8 leaf skills.

Frequently Asked Questions about autosearch:router

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

FAQPage Schema
How do I reduce context size when loading multiple skill files at session start?

Task routing minimizes context size by lazily loading only the matching leaf skills at runtime. It identifies 1–3 group indexes and selects 3–8 leaf skills from those groups, bypassing the need to load all files initially.

What is progressive-disclosure routing for task discovery?

Progressive-disclosure routing is a task discovery mechanism that narrows ambiguous queries by first selecting 1–3 group indexes, then extracting 3–8 leaf skills. This approach reduces startup overhead by loading only the most relevant skills on demand.

How does skill discovery handle ambiguous intent queries?

Skill discovery handles ambiguous intent by routing autosearch tasks to the smallest relevant group of skills. It selects 1–3 group indexes and 3–8 leaf skills, ensuring only the matching skills are lazily loaded.

Do I need mandatory frontmatter for skill discovery to work?

Yes, mandatory SKILL.md frontmatter with a name and description is required for skill discovery. This enforces standard metadata so the router can accurately identify and load the smallest relevant group of skills.

What happens if no group indexes match my task routing query?

When no group indexes match a task routing query, the system degrades gracefully. It falls back to generic web channels to ensure the autosearch task still receives a response without crashing.

Can I load optional resources like scripts and references on demand?

Yes, optional resources such as scripts, references, and assets are supported for on-demand loading. The router enforces mandatory SKILL.md frontmatter for discovery while allowing these supplementary resources to load only when needed.