Why is the SKILL.md file size limited?

Keeping the primary skill file small (under 500 lines) ensures agent efficiency. Detailed documentation is stored in reference files and loaded only when the agent needs specific context.

What is a NemoClaw Skill?

A NemoClaw skill is a markdown-based instruction module that teaches AI agents how to interact with tools and perform specialized tasks within a secured sandbox environment.

How do I install a new skill into a sandbox?

Skills can be mounted via the OpenShell policy YAML filesystem paths or manually copied into the ~/.openclaw/skills/ directory inside the connected sandbox.

What is the correct naming convention for skills?

Skill names should be lowercase, using only letters, numbers, and hyphens, and should ideally be written in gerund form (e.g., 'processing-data').

What does conditional gating do?

Conditional gating controls when a skill is active based on the availability of specific binaries, environment variables, or system configurations, preventing agents from attempting tasks without the necessary tools.

NemoClaw Skill Scaffolding

Name: NemoClaw Skill Scaffolding
Author: jawhnycooke

byjawhnycooke

0•

開発者ツール

Scaffolds custom OpenClaw AgentSkills for NemoClaw sandboxes using standardized templates and automated directory structures.

This skill streamlines the development of custom AI agent capabilities within NVIDIA NemoClaw sandboxed environments. By providing automated scaffolding for SKILL.md files and structured directory hierarchies, it enables developers to quickly author, test, and deploy domain-specific instructions. It features advanced conditional gating to ensure skills only load when necessary dependencies are met, and supports progressive disclosure patterns to keep agent context windows optimized for performance.

主な機能

01Built-in load precedence management for workspace and plugin skills

02Standardized directory structuring for reference docs, scripts, and templates

03Optimized progressive disclosure workflow for large documentation sets

040 GitHub stars

05Automated SKILL.md generation with required YAML frontmatter

06Environment-aware conditional gating for binary and config dependencies

ユースケース

01Implementing environment-locked skills that require specific CLI tools or API keys

02Authoring custom instruction sets for domain-specific sandboxed AI agents

03Standardizing agent skill development across large engineering teams

主な機能

01Built-in load precedence management for workspace and plugin skills

02Standardized directory structuring for reference docs, scripts, and templates

03Optimized progressive disclosure workflow for large documentation sets

040 GitHub stars

05Automated SKILL.md generation with required YAML frontmatter

06Environment-aware conditional gating for binary and config dependencies

ユースケース

01Implementing environment-locked skills that require specific CLI tools or API keys

02Authoring custom instruction sets for domain-specific sandboxed AI agents

03Standardizing agent skill development across large engineering teams