Skill Expert Skills

What it does

When your agent uses it

• : - Creating a new Skill (writing a SKILL.md) - Optimizing an existing Skill (structure
• Portability) - Validating a Skill package - Packaging
• Distributing a Skill Not for: regular programming
• Business logic (use domain-specific skills)

Example prompts

• “/skill-expert-skills”

Requirements

• Python 3
• Compatibility (from SKILL.md): Python 3.8+ for validation scripts
• Pre-approved tools (allowed-tools): Read, Write, Bash, Grep, Glob

Workflow steps

6 steps, taken from the step headings in SKILL.md.

1. 1Discovery + Hypothesis
2. 2Requirement Mining
3. 3Knowledge Acquisition
4. 4Skill Writing
5. 5Quality Validation + User Confirmation
6. 6Self-Reflection + Knowledge Precipitation

What it can do on your machine

Read from SKILL.md and the folder at commit e5199b5. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • Read
  • Write
  • Bash
  • Grep
  • Glob

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 1 file in scripts/, which the agent can run.

  Shell commands in SKILL.md call:

  • python

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Links to these hosts (documentation or services it may open):

  • platform.claude.com
  • agentskills.io
  • github.com

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

• Compatibility

  Python 3.8+ for validation scripts

  From compatibility in the SKILL.md frontmatter.

Context cost

Skill Expert Skills loads about 4.3k tokens when it runs, and up to ~111k if it reads all its reference files. Until then it costs about 170 tokens; SKILL.md has 1,203 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~170

When it runs · the whole SKILL.md, loaded when a task matches

~4.3k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~111k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• NotePre-approves every shell command (allowed-tools: Bash)SKILL.md

  allowed-tools: Read, Write, Bash, Grep, Glob

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from LeoYeAI/openclaw-master-skills at commit e5199b5, republished under its Apache-2.0 licence (© LeoYeAI). 1,203 words, ~4,320 tokens.

name

skill-expert-skills

description

Creates, optimizes, validates, and packages AI Agent Skills (SKILL.md format). Mandatory 6-Phase workflow with quality gates: Phase 0: Task Classification + Hypothesis Generation Phase 1: Deep Requirement Mining + 5 Whys Phase 2: Knowledge Acquisition + Validation Phase 3: Skill Writing + Quality Check Phase 4: Validation + User Confirmation Phase 5: Self-Reflection + Knowledge Precipitation Use when: - Creating a new Skill (writing a SKILL.md) - Optimizing an existing Skill (structure, triggers, portability) - Validating a Skill package - Packaging or distributing a Skill Not for: regular programming or business logic (use domain-specific skills).

allowed-tools

Read, Write, Bash, Grep, Glob

compatibility

Python 3.8+ for validation scripts

license

Apache-2.0

metadata.version

4.0.0

metadata.last_updated

2026-03-06

metadata.enhancement

v4.0 Added Fast Track Decision (Simple/Standard/Complex classification), v4.0 Adopted reference pointer pattern (-> references/xxx.md), v4.0 Added SKILL.md…

Task Classification
    |
    +-- Simple Skill (minimal template, < 100 lines, well-known domain)
    |   -> FAST TRACK: Phase 0 -> Phase 3 -> Phase 4
    |
    +-- Standard Skill (with references, 100-500 lines)
    |   -> STANDARD: Phase 0 -> Phase 1 -> Phase 2 -> Phase 3 -> Phase 4 -> Phase 5
    |
    +-- Complex Skill (knowledge-intensive, domain expertise needed)
    |   -> FULL: All phases with deep research
    |
    +-- Validate/Package Only
        -> Jump to Phase 4 / Command Reference

Stage 1: Explicit (5W1H)  ->  Stage 2: Implicit (4 methods)  ->  Stage 3: Validation

1) Comprehensive "summary"    2) Key-only "insight/diagnosis"
3) Produce "new content"      4) Reach a "conclusion"

LLM baseline -> Extract domains -> Research with tools -> Cross-validate -> Gate -> Self-check

SKILL.md SHOULD be:
  ✅ Scannable in 30 seconds (table of contents)
  ✅ Decision tree: "what situation → which action/file"
  ✅ Command reference: one-line key commands
  ✅ Minimal necessary constraints/contracts

SKILL.md should NOT be:
  ❌ Detailed knowledge base or tutorials
  ❌ Complete protocol explanations
  ❌ Long examples or code blocks
  ❌ Background knowledge

→ All detailed content MUST go to references/

---
name: my-skill              # Required. hyphen-case, ≤64 chars, matches directory name
description: |               # Required. ≤1024 chars, third person, no < >
  What this skill does.
  Use when:
  - scenario 1
  - scenario 2
  Not for: X, Y.
license: MIT                 # Optional
compatibility: Python 3.8+   # Optional. ≤500 chars
allowed-tools: Read Write    # Optional. space-delimited tool names
metadata:                    # Optional. extension fields
  version: 1.0.0
---

my-skill/
├── SKILL.md              # Required: instructions + metadata
├── scripts/              # Optional: executable code
│   ├── main.py
│   └── requirements.txt
├── references/           # Optional: detailed docs (loaded into context)
│   ├── patterns.md
│   └── checklist.md
└── assets/               # Optional: templates, images (NOT loaded into context)
    └── template.md

## Self-Reflection

| Dimension | Score (1-5) | Evidence |
|-----------|-------------|----------|
| Requirement Understanding | [1-5] | [notes] |
| Knowledge Completeness | [1-5] | [notes] |
| Output Quality | [1-5] | [notes] |
| User Satisfaction | [1-5] | [notes] |
| **Total** | **[/20]** | |

| Problem | Cause | Prevention |
|---------|-------|------------|
| [issue] | [why] | [measure] |

【Create New Skill】
  Phase 0: Classify task → Generate hypotheses → [Fast Track?] → User confirms
  Phase 1: 5 Whys → Skill Type → Validate requirements → User confirms
  Phase 2: Research domain → 4-Layer knowledge gate
  Phase 3: Select template → Write SKILL.md → Conciseness check
  Phase 4: Structural validation → Portability check → User confirms
  Phase 5: Self-reflect → Precipitate knowledge

【Optimize Existing Skill】
  Phase 0: Classify → Hypothesize what to improve → [Fast Track?] → User confirms
  Phase 1: 5 Whys on current pain points → User confirms
  Phase 2: Research latest patterns → 4-Layer gate
  Phase 3: Modify SKILL.md → Conciseness check
  Phase 4: Validate → User confirms
  Phase 5: Self-reflect → Document changes

【Validate / Package Only】
  -> Phase 4: Run validation scripts → Report results

# Search installed skills (reuse-first)
python scripts/search_skills.py "<keyword>" --root <skills-directory>

# Initialize new skill
python scripts/init_skill.py <skill-name> --path <skills-directory>

# Validate (required before delivery)
python scripts/quick_validate.py <skill-directory>
python scripts/universal_validate.py <skill-directory>

# Package for distribution (optional)
python scripts/package_skill.py <skill-directory> ./dist

# Maintenance
python scripts/upgrade_skill.py <skill-directory>
python scripts/diff_with_official.py <skill-directory>
python scripts/analyze_trigger.py <skill-directory>

© LeoYeAI, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 54 other files (scripts, references) in skills/skill-expert-skills-openclaw of LeoYeAI/openclaw-master-skills.

• SKILL.md
• LICENSE.txt
• QUICK_NAVIGATION.md
• _meta.json
• docs/_index.md
• references/deep-research-methodology.md
• references/docs-generation-workflow.md
• references/domain-expertise-protocol.md
• references/domain-knowledge-template.md
• references/domain-knowledge/_index.md
• references/domain-knowledge/backend-expertise.md
• references/domain-knowledge/bug-fixing-expertise.md
• references/domain-knowledge/code-review-expertise.md
• references/domain-knowledge/frontend-expertise.md
• references/examples.md
• references/hypothesis-ladder-for-skills.md
• references/integration-examples.md
• references/knowledge-acquisition-guide.md
• … and 37 more

Open the folder on GitHubat commit e5199b5