Skill Authoring

Create, edit, and package new AgentSkills. Use when asked to create a skill, build a skill, author a skill, or when需要对技能进行创建、打包或审核。Triggers on phrases like "...

Security Scan

The skill's files and runtime instructions are consistent with an authoring/packaging tool and do not request unrelated credentials or hidden network endpoints, but it references system paths and external helper scripts that you should verify before running.

Purpose & Capability

Name and description (Skill Authoring) align with the provided SKILL.md, references/design-patterns.md, and the helper script. The included list_skills.py and the content of SKILL.md are coherent with a workflow for creating, packaging, and inspecting AgentSkills.

Instruction Scope

SKILL.md instructs the user/agent to run external helper scripts (init_skill.py and package_skill.py) and to operate on system paths (examples reference /var/root/.openclaw/... and /usr/local/lib/node_modules/openclaw/...). Those actions are within the scope of authoring skills but require access to filesystem locations and external tooling that are not included in this package. No instructions attempt to read unrelated secrets or send data to external network endpoints.

Install Mechanism

There is no install spec — the skill is instruction-only plus a small utility script. That is low-risk. Note: the SKILL.md expects external tools (the skill-creator scripts) to exist in system locations; those tools are not part of this bundle and should be audited where they come from.

Credentials

The skill does not request environment variables, credentials, or config paths. The included script reads local skill directories (some of which are under /var/root), which is reasonable for a skill-management utility.

Persistence & Privilege

always is false and the skill does not request elevated or persistent privileges. It does not modify other skills' configurations in the files provided.

Assessment

This package appears to be a coherent skill-authoring guide plus a small utility to list installed skills. Before using it: 1) Verify the external helper scripts it tells you to run (init_skill.py and package_skill.py) are from a trusted source and inspect their code — they are not included here. 2) Avoid running commands as root or against /var/root unless you understand and control those locations. 3) Check any referenced system paths (/usr/local/lib/node_modules/openclaw/..., /var/root/.openclaw/...) exist and are owned by trusted software. 4) Review package_skill.py/init_skill.py for any network calls or data exfiltration before executing. If you cannot inspect the external tooling, treat running those scripts as higher risk.

Create modular, self-contained skills that extend agent capabilities.

The 6-Step Workflow

1. Understand → 2. Plan → 3. Init → 4. Edit → 5. Package → 6. Iterate

Step 1: Understand the Skill

Ask clarifying questions if needed. Understand:

• What functionality should this skill support?
• What are concrete usage examples?
• What should trigger this skill?

When to skip: When usage patterns are already clear.

Step 2: Plan Reusable Contents

Identify what to bundle:

• scripts/ — Executable code (Python/Bash) for deterministic tasks
• references/ — Documentation to load when needed
• assets/ — Files used in output (templates, icons)

Step 3: Initialize the Skill

Always use init_skill.py — never create manually:

python3 <skill-creator>/scripts/init_skill.py <name> --path <output-dir> --resources scripts,references,assets

Example:

python3 /usr/local/lib/node_modules/openclaw/skills/skill-creator/scripts/init_skill.py my-skill --path /var/root/.openclaw/workspace/skills --resources scripts,references

Step 4: Edit the Skill

SKILL.md Structure

---
name: skill-name
description: Clear description of what it does AND when to trigger it. Include Chinese triggers too.
---

# Skill Name

## Quick Start

## Details...

Naming Rules

• Lowercase, hyphens only
• Under 64 characters
• Verb-led preferred (e.g., pdf-edit, list-skills)

Key Principles

• Concise: Only add what the model doesn't already know
• Progressive disclosure: Keep SKILL.md lean, put details in references/
• Imperative mood: Use "do this" not "does this"

Step 5: Package the Skill

python3 <skill-creator>/scripts/package_skill.py <path/to/skill-folder>

Output: .skill file (zip format).

Validation runs automatically before packaging.

Step 6: Iterate

Test with real tasks, gather feedback, improve.

Skill Template

skill-name/
├── SKILL.md           # Required: name + description + instructions
├── scripts/           # Optional: executable code
├── references/        # Optional: load-as-needed docs
└── assets/            # Optional: output files

Quick Reference

Resource	When to Use
scripts/	Same code rewritten repeatedly
references/	Large docs the model should read on demand
assets/	Templates, images, boilerplate

Keep SKILL.md under 500 lines. Move details to references/.