Jeffrey's Skills Content Guidelines

This document defines the content requirements, restrictions, and quality standards for skills published on Jeffrey's Skills. These guidelines ensure a safe, useful, and high-quality skill marketplace.

Version: 1.0 Effective Date: January 2026

Table of Contents

1. Required Content
2. Prohibited Content
3. Quality Standards
4. Enforcement

1. Required Content

All skills must include the following components to be accepted for publication.

1.1 SKILL.md File (REQ-001)

Every skill package must contain a SKILL.md file at the root level with the following required sections:

1.2 Metadata (REQ-002)

Skills must include properly formatted metadata in the SKILL.md frontmatter:

---
name: my-skill-name
version: 1.0.0
author: Your Name
license: MIT
category: workflow
---

Required metadata fields:

• name - Unique identifier (lowercase, hyphens allowed)
• version - Semantic version (e.g., 1.0.0)
• author - Creator name or organization
• license - One of: MIT, Apache-2.0, GPL-3.0, BSD-3-Clause, proprietary, all-rights-reserved
• category - One of: workflow, code-generation, code-review, testing, debugging, documentation, refactoring, architecture, devops, brainstorming, learning, productivity, flywheel-tool, multi-agent, other

1.3 Attribution (REQ-003)

• Original authors must be credited in the author field
• Derivative works must acknowledge the original skill
• Third-party code snippets must include appropriate license notices

1.4 Distribution Declaration (REQ-004)

Skills must declare their distribution rule:

• public - Available to all users
• subscribers - Available to paying subscribers only
• forbidden - Not available for distribution (private use only)

2. Prohibited Content

The following content types are strictly prohibited and will result in skill rejection or removal.

2.1 Protected Names and Terms (PRO-001)

The following terms and prefixes are reserved and cannot be used in skill names:

Protected Terms:

• writing-skills, writing-skill
• skill-creation, skill-creator
• skill-writer, meta-skill
• creating-skills

Protected Prefixes:

• jeffrey- (reserved for official skills)
• vault- (reserved for system skills)
• internal- (reserved for internal use)

Note: Attempts to evade these restrictions using l33t speak (e.g., wr1ting-sk1lls, j3ffr3y-) or similar substitutions will be detected and rejected.

2.2 Executable and Binary Files (PRO-002)

Skills cannot contain executable or binary files:

• .exe - Windows executables
• .dll - Windows dynamic libraries
• .so - Linux shared objects
• .dylib - macOS dynamic libraries

2.3 Secrets and Credentials (PRO-003)

Skills must not contain hardcoded secrets, credentials, or authentication tokens. The following patterns are scanned and rejected:

• API keys (api_key, apikey, api-key)
• Secret keys (secret_key, secretkey, secret-key)
• Passwords (password, passwd)
• Private keys (private_key, privatekey)
• Access tokens (access_token, bearer )

Exception: Placeholder values clearly marked as examples (e.g., YOUR_API_KEY_HERE) are permitted.

2.4 URL Shorteners (PRO-004)

Skills must not contain shortened URLs. Use full, transparent URLs instead:

Prohibited shorteners:

• bit.ly, tinyurl.com, t.co, goo.gl, ow.ly
• is.gd, buff.ly, rebrand.ly, short.io, cutt.ly

2.5 Meta-Skill Content (PRO-005)

Skills that teach how to create, write, or generate other skills are prohibited. This includes content matching patterns such as:

• "how to create a skill"
• "how to write a skill"
• "skill authoring guide"
• "skill creation tutorial"
• "skill generator"
• "create your own skill"
• "build a skill"
• "skill template generator"

2.6 Malicious Content (PRO-006)

The following are strictly prohibited:

• Malware, viruses, or code designed to harm systems
• Phishing content or credential harvesting
• Exploitation tools or vulnerability scanners intended for malicious use
• Denial of service tools
• Cryptocurrency mining without explicit user consent

2.7 Illegal Content (PRO-007)

Skills must not contain or facilitate:

• Content that violates applicable laws
• Pirated software or circumvention tools
• Content that infringes intellectual property rights
• Instructions for illegal activities

2.8 Harassment and Harmful Content (PRO-008)

Skills must not contain:

• Hate speech or discriminatory content
• Harassment or threats
• Personal information without consent (doxxing)
• Content that promotes self-harm
• Explicit adult content without appropriate labeling

3. Quality Standards

Skills must meet the following quality standards to be published.

3.1 Description Quality (QUA-001)

• Minimum length: 50 characters
• Clarity: Description must clearly explain what the skill does
• Accuracy: Description must accurately reflect the skill's functionality
• Grammar: Reasonable grammar and spelling expected

3.2 Size Limits (QUA-002)

3.3 File Organization (QUA-003)

• SKILL.md must be at the package root
• File names should be descriptive and use lowercase with hyphens
• Directory structure should be logical and navigable
• No deeply nested directories (max 5 levels recommended)

3.4 Usefulness (QUA-004)

Skills should:

• Solve a real problem or provide genuine value
• Be well-tested and functional
• Include clear usage examples in the description
• Not duplicate existing official skills without significant improvement

3.5 Trigger Clarity (QUA-005)

Triggers must:

• Clearly specify when the skill should activate
• Not conflict with common user commands
• Be specific enough to avoid false activations
• Be documented in the Triggers section

4. Enforcement

4.1 Review Process

All skills undergo automated validation checks followed by optional human review for edge cases.

Automated checks include:

• Metadata validation (REQ-002)
• Protected name detection (PRO-001)
• File type scanning (PRO-002)
• Secrets detection (PRO-003)
• URL shortener detection (PRO-004)
• Size limit validation (QUA-002)

4.2 Violation Consequences

4.3 Appeals

Authors may appeal moderation decisions by contacting support with:

• The skill ID or name
• The guideline section cited
• Explanation of why the decision should be reconsidered

4.4 Reporting Violations

Users can report skills that violate these guidelines through:

• The "Report" button on skill pages
• Email to the moderation team
• Community Discord reporting channel

Guideline Reference Codes

For moderator use when citing violations:

Changelog

• 1.0 (January 2026): Initial release