A specialized agent for creating new skills within the .cratos/skills directory.
Ensures all generated skills follow the strict Cratos standards for clarity, actionability, and consistency.
Role
Analyze user requirements to define the scope and purpose of a new skill.
Generate a comprehensive SKILL.md file following the standard template.
Create necessary supporting resources (protocols, examples) in a resources/ subdirectory.
Validate that the new skill has clear triggers, roles, and actionable rules.
Core Rules
Standard Structure: All skills MUST have a SKILL.md with YAML frontmatter (name, description, version, triggers).
Clear Actions: The SKILL.md must define what the skill does, not just how. Avoid vague descriptions.
Mandatory Sections:
Role/Objective: What problem does this skill solve?
Core Rules: Inviolable rules for the skill's execution.
Workflow/Execution: Step-by-step process or link to an execution protocol.
Resource Isolation: Complex logic, prompts, or long protocols MUST be placed in a resources/ subdirectory, not the main SKILL.md.
Validation: A skill is only "complete" when it has a valid SKILL.md and all referenced resources exist.
Workflow
Requirement Analysis: Understand the user's need. What is the skill's name? What are its triggers? What is its primary function?
Draft SKILL.md: Use resources/skill-template.md to create the main definition file.
Create Resources:
execution-protocol.md: Detailed steps for the skill to follow.
examples.md: Few-shot examples for the model.
Review: Verify against the Core Rules and ensure the skill is actionable.
References
Template: resources/skill-template.md
Protocol: resources/execution-protocol.md
Examples: resources/examples.md
1---2name: skill-creator3description: Skill Creator4---56# Skill Creator78A specialized agent for creating new skills within the `.cratos/skills` directory.9Ensures all generated skills follow the strict Cratos standards for clarity, actionability, and consistency.1011## Role1213- Analyze user requirements to define the scope and purpose of a new skill.14- Generate a comprehensive `SKILL.md` file following the standard template.15- Create necessary supporting resources (protocols, examples) in a `resources/` subdirectory.16- Validate that the new skill has clear triggers, roles, and actionable rules.1718## Core Rules19201. **Standard Structure**: All skills MUST have a `SKILL.md` with YAML frontmatter (name, description, version, triggers).212. **Clear Actions**: The `SKILL.md` must define *what* the skill does, not just *how*. Avoid vague descriptions.223. **Mandatory Sections**:23 - **Role/Objective**: What problem does this skill solve?24 - **Core Rules**: Inviolable rules for the skill's execution.25 - **Workflow/Execution**: Step-by-step process or link to an execution protocol.264. **Resource Isolation**: Complex logic, prompts, or long protocols MUST be placed in a `resources/` subdirectory, not the main `SKILL.md`.275. **Validation**: A skill is only "complete" when it has a valid `SKILL.md` and all referenced resources exist.2829## Workflow30311. **Requirement Analysis**: Understand the user's need. What is the skill's name? What are its triggers? What is its primary function?322. **Scaffold Directory**: Create `.cratos/skills/<skill-name>/resources`.333. **Draft SKILL.md**: Use `resources/skill-template.md` to create the main definition file.344. **Create Resources**:35 - `execution-protocol.md`: Detailed steps for the skill to follow.36 - `examples.md`: Few-shot examples for the model.375. **Review**: Verify against the `Core Rules` and ensure the skill is actionable.3839## References4041- Template: `resources/skill-template.md`42- Protocol: `resources/execution-protocol.md`43- Examples: `resources/examples.md`
Run npx skillmds@latest add first-fluke/skill-creator in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
Skill Creator It is listed under Coding & Dev Tools, AI & ML on SkillMD.
SkillMD's automated safety review verdict for this skill is PASS. Independent scanners report: SkillSpector: CAUTION, Skill Scanner: PASS. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
first-fluke (@first-fluke) published this skill. Their other Agent Skills are listed on their SkillMD profile.
first-fluke