Agent-Skills.md

Agent Skills: Toolkit Registry Standards

Provides System Authority on 2026 Universal Agentic Runtime standards for all plugin components. MUST Use when auditing or managing plugin components (Skills, Commands, Agents). Do not use for creating new components, development tasks, or routine maintenance.

UncategorizedID: Git-Fg/thecattoolkit/adhering-standards

Repository

Git-Fg/thecattoolkit
Git-FgLicense: MIT
1

Install this agent skill to your local

pnpm dlx add-skill https://github.com/Git-Fg/thecattoolkit/tree/HEAD/plugins/sys-core/skills/adhering-standards

Skill Files

Browse the full folder contents for adhering-standards.

Download Skill

Loading file tree…

plugins/sys-core/skills/adhering-standards/SKILL.md

Skill Metadata

Name
adhering-standards
Description
"Provides System Authority on 2026 Universal Agentic Runtime standards for all plugin components. MUST Use when auditing or managing plugin components (Skills, Commands, Agents). Do not use for creating new components, development tasks, or routine maintenance."

Toolkit Registry Standards

Quick Reference

| Component | Purpose | Entry Trigger | |:----------|:--------|:--------------| | Skills | Knowledge injection, reusable workflows | "Create a skill for X", "Audit skill Y" | | Commands | Orchestration shortcuts, multi-skill workflows | "Create a command for X", "Audit command Y" | | Agents | Persona binding, isolated context execution | "Create an agent for X", "Audit agent Y" |

Core Standards

Component Management

For detailed specifications, see: references/full-spec.md

Key Standards:

  • Skills: USE when patterns, allowed-tools, progressive disclosure
  • Commands: disable-model-invocation, argument-hint, orchestration focus
  • Agents: Persona binding, tool restrictions, no AskUser in workers

Validation Protocol

All components MUST pass validation before use. See references/validation.md for complete validation protocol.