Grok Custom Skills
Description: Central repository for reusable, safety-conscious agent skills compatible with Grok and other AI agent frameworks. A curated collection of 89 skill packages spanning workflow automation, research, integrations, quality assurance, and more.
Purpose: Provide a comprehensive, well-documented, and safe library of skills that enable AI agents to perform complex tasks while maintaining transparency, control, and safety.
🎯 Quick Start
For AI Agent Developers
- Browse available skills in SKILLS_INDEX.md
- Clone the repository
- Copy the SKILL.md files you need to your agent's skills directory
- Configure any required settings or credentials
- Test with sandbox environments first
- Deploy and monitor usage
# Clone the repository
git clone https://github.com/Stijnman/grok-custom-skills.git
cd grok-custom-skills
# Install the complete collection
mkdir -p ~/.grok/skills
cp -a .grok/skills/. ~/.grok/skills/
# Or install one specific skill
cp -a .grok/skills/drive-github-skill-audit ~/.grok/skills/
For Users
Tell your AI agent:
- "Research the latest AI trends from reputable sources"
- "Create a comprehensive project plan with milestones"
- "Audit this GitHub repository for MCP security issues"
- "Generate an accessible color palette for my website"
- "Plan a multi-step workflow for my task"
Your agent will use the appropriate skill based on your request.
📦 Skill Categories
The repository is organized into 7 main categories with 89 total skill packages:
| Category |
Count |
Description |
Example Skills |
| Workflow & Orchestration |
15 |
Plan multi-step work, coordinate agents, handle failures, verify goals |
goal-verifier, workflow-orchestrator |
| Skill Development & Operations |
12 |
Create, audit, evolve, test, and compare skill collections |
skill-creator, skill-rubric-reviewer |
| Safety, Privacy & Governance |
10 |
Require human approval, redact data, audit MCP exposure, responsible browsing |
hitl-approver, privacy-redactor, mcp-auditor |
| Research, Web & Integrations |
20 |
Perform source-aware research, inspect repositories, discover tools, connect services |
deep-search-enabler, repository-inspector, tool-discoverer |
| Memory, Context & Knowledge |
12 |
Manage context budgets, structured knowledge, semantic memory, session handoffs |
context-budget-manager, semantic-memory, knowledge-structurer |
| Media, Voice & Visual Work |
10 |
Generate or edit images, analyze video, work with voice, prepare song-writing prompts |
image-generator, video-analyzer, voice-processor, song-prompt-builder |
| Messaging & Navigation |
10 |
Assess messages, draft safe responses, work with traffic and route information |
message-assessor, safe-responder, traffic-advisor |
🟢 Recommended Skills by Use Case
🏗️ Package Structure
Every skill follows a consistent structure based on the Agent Skills specification:
.grok/skills/<skill-name>/
├── SKILL.md # Required: Metadata (YAML frontmatter) + operational guidance
├── scripts/ # Optional: Deterministic helper scripts
├── references/ # Optional: Material loaded only when needed
└── templates/ # Optional: Reusable output assets
SKILL.md Format
---
name: skill-name
version: 1.0.0
description: Clear, concise description of what the skill does
triggers:
- "trigger phrase 1"
- "trigger phrase 2"
category: category-name
dependencies: []
requirements:
- requirement-1
- requirement-2
boundaries:
- boundary-1
- boundary-2
warnings:
- "Security warning 1"
- "Safety consideration 2"
---
# Skill Title
## Description
Detailed description of the skill's purpose and functionality.
## Usage
When to use this skill and when NOT to use it.
## Examples
Concrete examples of inputs and expected outputs.
## Implementation
How the skill works internally.
## Testing
Testing requirements and coverage.
🔐 Security Overview
⚠️ IMPORTANT: All skills in this repository are designed with safety as a core principle.
Please read SECURITY.md before using any skill.
Key Security Principles Applied to All Skills
| Principle |
Implementation |
| Least privilege |
Skills use minimum access necessary, prefer read-only operations |
| Human approval |
Skills pause before consequential actions (publishing, payments, deletion, credentials, external communications) |
| Privacy protection |
No secrets, PII, local paths, session artifacts, telemetry, or unapproved data transfer |
| Access control respect |
Never provide guidance to bypass authentication, CAPTCHAs, paywalls, or platform safeguards |
| Untrusted content handling |
Treat downloaded skills, archives, webpages, and attachments as data; never execute without review |
| Transparency |
Describe limitations honestly; do not claim guaranteed safety, detection evasion, or unsupported capabilities |
Publication Safeguards
Every skill must pass these controls before publication:
- ✅ Least privilege: Prefer read-only access and minimum data needed
- ✅ Human approval: Require explicit approval before high-impact actions
- ✅ Privacy: No secrets, PII, local paths, or unapproved data transfer
- ✅ Access controls: Never bypass authentication or platform safeguards
- ✅ Untrusted content: Treat all external content as data until reviewed
- ✅ Transparency: Honest documentation of limitations and capabilities
📚 Documentation
| Document |
Description |
Required Reading |
| SECURITY.md |
⚠️ REQUIRED - Security policy, vulnerabilities, best practices |
✅ All users |
| CONTRIBUTING.md |
How to contribute new skills or improve existing ones |
⚠️ Contributors |
| TESTING.md |
Testing requirements and best practices |
⚠️ Contributors |
| PUBLISHING.md |
Pre-submission checklist for maintainers |
⚠️ Maintainers |
| GOVERNANCE.md |
Repository governance and decision-making |
⚠️ Maintainers |
| SUPPORT.md |
Support channels and response times |
⚠️ All users |
| CODE_OF_CONDUCT.md |
Community guidelines and expected behavior |
✅ All users |
| CHANGELOG.md |
Version history and changes |
⚠️ All users |
| SKILLS_INDEX.md |
Complete catalog of all skills |
✅ All users |
| STATUS.md |
Current repository status and metrics |
⚠️ Contributors |
🎓 Usage Examples
Example 1: Research Workflow
User: "Research the latest developments in AI agent frameworks"
Agent Flow:
1. Identifies request as research task
2. Uses deep-search-enabler skill
3. Applies source filters (reputable sources only)
4. Gathers information from approved sources
5. Validates information quality
6. Returns: Structured research results with sources
7. Cites: All sources used
Example 2: GitHub Repository Audit
User: "Audit this GitHub repository for MCP security issues"
Agent Flow:
1. Identifies request as MCP audit
2. Uses defensive-mcp-audit skill
3. Clones or accesses the repository
4. Scans for risky localhost exposure
5. Checks for weak bindings
6. Identifies confused-deputy risks
7. Returns: SARIF + HTML report of findings
Example 3: Multi-Step Workflow
User: "Plan a comprehensive product launch strategy"
Agent Flow:
1. Uses product-opportunity-tree to map opportunities
2. Uses product-requirements-document to define requirements
3. Uses architecture-decision-record to document decisions
4. Uses workflow-orchestrator to coordinate steps
5. Uses goal-verifier to validate completion
6. Returns: Complete product launch plan with milestones
Example 4: Skill Development
User: "Create a new skill for web scraping"
Agent Flow:
1. Uses skill-creator to generate template
2. Uses skill-rubric-reviewer to check quality
3. Validates against security requirements
4. Tests with sandbox inputs
5. Updates SKILLS_INDEX.md
6. Returns: New skill package ready for review
🧪 Testing
All skills have been tested for production use. See TESTING.md for:
- Testing philosophy and requirements
- Manual testing checklists
- Automated testing setup
- Skill category-specific testing guides
- CI/CD pipeline configuration
Quality Assurance:
- ✅ All skills pass validation pipeline
- ✅ All skills have security checks
- ✅ All skills have documented boundaries
- ✅ Catalog is auto-generated from live metadata
- ✅ All publication checks pass
🤝 Contributing
We welcome contributions! Please read CONTRIBUTING.md for:
- How to add new skills
- How to improve existing skills
- Testing requirements
- Pull request process
- Code of conduct
Quick Contribution Guide
- Fork the repository
- Clone your fork
- Create a feature branch (
git checkout -b feat/amazing-skill)
- Read PUBLISHING.md checklist
- Run validation:
python3 scripts/optimize_all_skills.py
- Run tests:
python3 -m unittest discover -s tests -p "test_*.py"
- Commit with clear messages
- Push to your fork
- Open a Pull Request
📜 License
This repository is licensed under the MIT License. See LICENSE for full license text.
You are free to:
- Use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
- Use for commercial purposes
- Modify for your own needs
Under the following conditions:
- Include copyright notice and license in all copies
- Provide attribution to the original author (Stijnman)
Note: Individual skill packages may have different licenses (e.g., Apache-2.0 for product-management workflows with required attribution in THIRD_PARTY_NOTICES.md).
📞 Support & Contact
🏷️ Repository Metadata
| Attribute |
Value |
| Repository |
grok-custom-skills |
| Owner |
Stijnman |
| License |
MIT |
| Total Skills |
89 |
| Categories |
7 |
| Language |
English |
| Created |
July 2026 |
| Last Updated |
September 2026 |
| Stars |
9 |
| Forks |
1 |
Categories Breakdown
| Category |
Skills |
Percentage |
| Research & Integrations |
20 |
22.5% |
| Workflow & Orchestration |
15 |
16.9% |
| Skill Development |
12 |
13.5% |
| Memory & Knowledge |
12 |
13.5% |
| Media & Visual |
10 |
11.2% |
| Messaging & Navigation |
10 |
11.2% |
| Safety & Governance |
10 |
11.2% |
🔗 Related Resources
Specifications & Standards
Tools & Frameworks
Documentation
🎯 Key Differentiators
| Feature |
This Repository |
Other Skill Repositories |
| Safety Focus |
✅ Core principle - all skills have safety checks |
❌ Varies by contributor |
| Consistent Structure |
✅ All skills follow same format |
❌ Inconsistent formats |
| Comprehensive Testing |
✅ Testing guide and requirements |
❌ Minimal testing |
| Publication Safeguards |
✅ 6 mandatory controls |
❌ No controls |
| Auto-Generated Catalog |
✅ SKILLS_INDEX.md from live metadata |
❌ Manual catalog |
| Validation Pipeline |
✅ Automated checks |
❌ No validation |
| Documentation Quality |
✅ 10/10 standard |
❌ Varies widely |
📌 Important Notes
Design Principle
Skills should make agent behavior more useful, auditable, and safe. They are guidance packages, not permission to bypass controls, disclose private data, or automate high-impact actions without human approval.
Usage Warning
Always read and understand a skill before using it. Check the triggers, boundaries, and warnings in each SKILL.md file. Follow all security requirements in SECURITY.md.
Contribution Warning
Do not submit skills that: bypass authentication, evade CAPTCHAs, circumvent paywalls, harvest credentials, automate spam, violate platform terms, or claim unsafe capabilities.
🏆 Version History
| Date |
Version |
Changes |
| 2026-09-11 |
2.0.0 |
Added comprehensive documentation (SKILL.md, TESTING.md, STATUS.md, CONTRIBUTORS.md), pre-commit hooks, markdown lint config, secrets baseline |
| 2026-09-10 |
1.2.0 |
Enhanced documentation (SECURITY.md, CONTRIBUTING.md, CODE_OF_CONDUCT.md, CHANGELOG.md, GOVERNANCE.md, PUBLISHING.md, SUPPORT.md), added .github workflows and templates |
| 2026-07-30 |
1.0.0 |
Initial repository with 89 skill packages, catalog system, validation pipeline |
Last updated: September 11, 2026
Maintainer: Stijnman
Repository: grok-custom-skills
Need help? See SUPPORT.md or open a GitHub issue.
1---2name: grok-custom-skills3description: Grok Custom Skills4metadata:5 source: https://skillmd.com/skills/stijnman/grok-custom-skills6---7# Grok Custom Skills89**Description**: Central repository for reusable, safety-conscious agent skills compatible with Grok and other AI agent frameworks. A curated collection of 89 skill packages spanning workflow automation, research, integrations, quality assurance, and more.1011**Purpose**: Provide a comprehensive, well-documented, and safe library of skills that enable AI agents to perform complex tasks while maintaining transparency, control, and safety.1213---1415## 🎯 Quick Start1617### For AI Agent Developers18191. **Browse available skills** in [SKILLS_INDEX.md](./SKILLS_INDEX.md)202. **Clone** the repository213. **Copy** the SKILL.md files you need to your agent's skills directory224. **Configure** any required settings or credentials235. **Test** with sandbox environments first246. **Deploy** and monitor usage2526```bash27# Clone the repository28git clone https://github.com/Stijnman/grok-custom-skills.git29cd grok-custom-skills3031# Install the complete collection32mkdir -p ~/.grok/skills33cp -a .grok/skills/. ~/.grok/skills/3435# Or install one specific skill36cp -a .grok/skills/drive-github-skill-audit ~/.grok/skills/37```3839### For Users4041Tell your AI agent:42- "Research the latest AI trends from reputable sources"43- "Create a comprehensive project plan with milestones"44- "Audit this GitHub repository for MCP security issues"45- "Generate an accessible color palette for my website"46- "Plan a multi-step workflow for my task"4748Your agent will use the appropriate skill based on your request.4950---5152## 📦 Skill Categories5354The repository is organized into **7 main categories** with **89 total skill packages**:5556| Category | Count | Description | Example Skills |57|----------|-------|-------------|---------------|58| **Workflow & Orchestration** | 15 | Plan multi-step work, coordinate agents, handle failures, verify goals | goal-verifier, workflow-orchestrator |59| **Skill Development & Operations** | 12 | Create, audit, evolve, test, and compare skill collections | skill-creator, skill-rubric-reviewer |60| **Safety, Privacy & Governance** | 10 | Require human approval, redact data, audit MCP exposure, responsible browsing | hitl-approver, privacy-redactor, mcp-auditor |61| **Research, Web & Integrations** | 20 | Perform source-aware research, inspect repositories, discover tools, connect services | deep-search-enabler, repository-inspector, tool-discoverer |62| **Memory, Context & Knowledge** | 12 | Manage context budgets, structured knowledge, semantic memory, session handoffs | context-budget-manager, semantic-memory, knowledge-structurer |63| **Media, Voice & Visual Work** | 10 | Generate or edit images, analyze video, work with voice, prepare song-writing prompts | image-generator, video-analyzer, voice-processor, song-prompt-builder |64| **Messaging & Navigation** | 10 | Assess messages, draft safe responses, work with traffic and route information | message-assessor, safe-responder, traffic-advisor |6566---6768## 🟢 Recommended Skills by Use Case6970| Goal | Start with | Category |71|------|-----------|----------|72| **Compare a Drive skill library with GitHub** | [`drive-github-skill-audit`](./.grok/skills/drive-github-skill-audit/SKILL.md) | Research & Integrations |73| **Create or improve a skill package** | [`skill-creator`](./.grok/skills/skill-creator/SKILL.md) and [`skill-rubric-reviewer`](./.grok/skills/skill-rubric-reviewer/SKILL.md) | Skill Development |74| **Verify a completed task** | [`goal-verifier`](./.grok/skills/goal-verifier/SKILL.md) | Workflow & Orchestration |75| **Review code changes** | [`code-reviewer`](./.grok/skills/code-reviewer/SKILL.md) and [`auto-tester`](./.grok/skills/auto-tester/SKILL.md) | Skill Development |76| **Research public information responsibly** | [`deep-search-enabler`](./.grok/skills/deep-search-enabler/SKILL.md) and [`sandbox-internet-handler`](./.grok/skills/sandbox-internet-handler/SKILL.md) | Research & Integrations |77| **Handle high-impact actions** | [`hitl-approver`](./.grok/skills/hitl-approver/SKILL.md) | Safety & Governance |78| **Structure product discovery and delivery** | [`product-opportunity-tree`](./.grok/skills/product-opportunity-tree/SKILL.md), [`product-requirements-document`](./.grok/skills/product-requirements-document/SKILL.md) | Workflow & Orchestration |79| **Prepare a safe contact import file** | [`contact-vcard-export`](./.grok/skills/contact-vcard-export/SKILL.md) | Messaging & Navigation |80| **Review visual accessibility** | [`accessible-color-review`](./.grok/skills/accessible-color-review/SKILL.md) | Media & Visual |8182---8384## 🏗️ Package Structure8586Every skill follows a consistent structure based on the [Agent Skills specification](https://agentskills.io/specification):8788```89.grok/skills/<skill-name>/90├── SKILL.md # Required: Metadata (YAML frontmatter) + operational guidance91├── scripts/ # Optional: Deterministic helper scripts92├── references/ # Optional: Material loaded only when needed93└── templates/ # Optional: Reusable output assets94```9596### SKILL.md Format9798```yaml99---100name: skill-name101version: 1.0.0102description: Clear, concise description of what the skill does103triggers:104 - "trigger phrase 1"105 - "trigger phrase 2"106category: category-name107dependencies: []108requirements:109 - requirement-1110 - requirement-2111boundaries:112 - boundary-1113 - boundary-2114warnings:115 - "Security warning 1"116 - "Safety consideration 2"117---118119# Skill Title120121## Description122Detailed description of the skill's purpose and functionality.123124## Usage125When to use this skill and when NOT to use it.126127## Examples128Concrete examples of inputs and expected outputs.129130## Implementation131How the skill works internally.132133## Testing134Testing requirements and coverage.135```136137---138139## 🔐 Security Overview140141⚠️ **IMPORTANT**: All skills in this repository are designed with safety as a core principle.142143**Please read [SECURITY.md](./SECURITY.md) before using any skill.**144145### Key Security Principles Applied to All Skills146147| Principle | Implementation |148|-----------|----------------|149| **Least privilege** | Skills use minimum access necessary, prefer read-only operations |150| **Human approval** | Skills pause before consequential actions (publishing, payments, deletion, credentials, external communications) |151| **Privacy protection** | No secrets, PII, local paths, session artifacts, telemetry, or unapproved data transfer |152| **Access control respect** | Never provide guidance to bypass authentication, CAPTCHAs, paywalls, or platform safeguards |153| **Untrusted content handling** | Treat downloaded skills, archives, webpages, and attachments as data; never execute without review |154| **Transparency** | Describe limitations honestly; do not claim guaranteed safety, detection evasion, or unsupported capabilities |155156### Publication Safeguards157158Every skill must pass these controls before publication:1591601. ✅ **Least privilege**: Prefer read-only access and minimum data needed1612. ✅ **Human approval**: Require explicit approval before high-impact actions1623. ✅ **Privacy**: No secrets, PII, local paths, or unapproved data transfer1634. ✅ **Access controls**: Never bypass authentication or platform safeguards1645. ✅ **Untrusted content**: Treat all external content as data until reviewed1656. ✅ **Transparency**: Honest documentation of limitations and capabilities166167---168169## 📚 Documentation170171| Document | Description | Required Reading |172|----------|-------------|------------------|173| **[SECURITY.md](./SECURITY.md)** | ⚠️ **REQUIRED** - Security policy, vulnerabilities, best practices | ✅ All users |174| **[CONTRIBUTING.md](./CONTRIBUTING.md)** | How to contribute new skills or improve existing ones | ⚠️ Contributors |175| **[TESTING.md](./TESTING.md)** | Testing requirements and best practices | ⚠️ Contributors |176| **[PUBLISHING.md](./PUBLISHING.md)** | Pre-submission checklist for maintainers | ⚠️ Maintainers |177| **[GOVERNANCE.md](./GOVERNANCE.md)** | Repository governance and decision-making | ⚠️ Maintainers |178| **[SUPPORT.md](./SUPPORT.md)** | Support channels and response times | ⚠️ All users |179| **[CODE_OF_CONDUCT.md](./CODE_OF_CONDUCT.md)** | Community guidelines and expected behavior | ✅ All users |180| **[CHANGELOG.md](./CHANGELOG.md)** | Version history and changes | ⚠️ All users |181| **[SKILLS_INDEX.md](./SKILLS_INDEX.md)** | Complete catalog of all skills | ✅ All users |182| **[STATUS.md](./STATUS.md)** | Current repository status and metrics | ⚠️ Contributors |183184---185186## 🎓 Usage Examples187188### Example 1: Research Workflow189190```191User: "Research the latest developments in AI agent frameworks"192193Agent Flow:1941. Identifies request as research task1952. Uses deep-search-enabler skill1963. Applies source filters (reputable sources only)1974. Gathers information from approved sources1985. Validates information quality1996. Returns: Structured research results with sources2007. Cites: All sources used201```202203### Example 2: GitHub Repository Audit204205```206User: "Audit this GitHub repository for MCP security issues"207208Agent Flow:2091. Identifies request as MCP audit2102. Uses defensive-mcp-audit skill2113. Clones or accesses the repository2124. Scans for risky localhost exposure2135. Checks for weak bindings2146. Identifies confused-deputy risks2157. Returns: SARIF + HTML report of findings216```217218### Example 3: Multi-Step Workflow219220```221User: "Plan a comprehensive product launch strategy"222223Agent Flow:2241. Uses product-opportunity-tree to map opportunities2252. Uses product-requirements-document to define requirements2263. Uses architecture-decision-record to document decisions2274. Uses workflow-orchestrator to coordinate steps2285. Uses goal-verifier to validate completion2296. Returns: Complete product launch plan with milestones230```231232### Example 4: Skill Development233234```235User: "Create a new skill for web scraping"236237Agent Flow:2381. Uses skill-creator to generate template2392. Uses skill-rubric-reviewer to check quality2403. Validates against security requirements2414. Tests with sandbox inputs2425. Updates SKILLS_INDEX.md2436. Returns: New skill package ready for review244```245246---247248## 🧪 Testing249250All skills have been tested for production use. See [TESTING.md](./TESTING.md) for:251252- Testing philosophy and requirements253- Manual testing checklists254- Automated testing setup255- Skill category-specific testing guides256- CI/CD pipeline configuration257258**Quality Assurance**:259- ✅ All skills pass validation pipeline260- ✅ All skills have security checks261- ✅ All skills have documented boundaries262- ✅ Catalog is auto-generated from live metadata263- ✅ All publication checks pass264265---266267## 🤝 Contributing268269We welcome contributions! Please read [CONTRIBUTING.md](./CONTRIBUTING.md) for:270271- How to add new skills272- How to improve existing skills273- Testing requirements274- Pull request process275- Code of conduct276277### Quick Contribution Guide2782791. **Fork** the repository2802. **Clone** your fork2813. **Create** a feature branch (`git checkout -b feat/amazing-skill`)2824. **Read** [PUBLISHING.md](./PUBLISHING.md) checklist2835. **Run** validation: `python3 scripts/optimize_all_skills.py`2846. **Run** tests: `python3 -m unittest discover -s tests -p "test_*.py"`2857. **Commit** with clear messages2868. **Push** to your fork2879. **Open** a Pull Request288289---290291## 📜 License292293This repository is licensed under the **MIT License**. See [LICENSE](./LICENSE) for full license text.294295**You are free to**:296- Use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies297- Use for commercial purposes298- Modify for your own needs299300**Under the following conditions**:301- Include copyright notice and license in all copies302- Provide attribution to the original author (Stijnman)303304**Note**: Individual skill packages may have different licenses (e.g., Apache-2.0 for product-management workflows with required attribution in THIRD_PARTY_NOTICES.md).305306---307308## 📞 Support & Contact309310| Issue Type | How to Get Help | Response Time | Notes |311|-----------|-----------------|---------------|-------|312| 🐛 **Bug Report** | [Open a GitHub Issue](https://github.com/Stijnman/grok-custom-skills/issues/new?template=bug_report.yml) | 24-48 hours | Use bug_report.yml template |313| 🔒 **Security Issue** | Private message via GitHub profile | Immediate | See [SECURITY.md](./SECURITY.md) |314| ❓ **General Question** | [Open a GitHub Discussion](https://github.com/Stijnman/grok-custom-skills/discussions) | 24 hours | For non-sensitive questions |315| 💡 **Feature Request** | [Open a GitHub Issue](https://github.com/Stijnman/grok-custom-skills/issues/new?template=feature_request.yml) | 1 week | Use feature_request.yml template |316317---318319## 🏷️ Repository Metadata320321| Attribute | Value |322|-----------|-------|323| **Repository** | [grok-custom-skills](https://github.com/Stijnman/grok-custom-skills) |324| **Owner** | [Stijnman](https://github.com/Stijnman) |325| **License** | MIT |326| **Total Skills** | 89 |327| **Categories** | 7 |328| **Language** | English |329| **Created** | July 2026 |330| **Last Updated** | September 2026 |331| **Stars** | 9 |332| **Forks** | 1 |333334### Categories Breakdown335336| Category | Skills | Percentage |337|----------|--------|------------|338| Research & Integrations | 20 | 22.5% |339| Workflow & Orchestration | 15 | 16.9% |340| Skill Development | 12 | 13.5% |341| Memory & Knowledge | 12 | 13.5% |342| Media & Visual | 10 | 11.2% |343| Messaging & Navigation | 10 | 11.2% |344| Safety & Governance | 10 | 11.2% |345346---347348## 🔗 Related Resources349350### Specifications & Standards351- [Agent Skills specification](https://agentskills.io/specification) - Skill package conventions352- [Model Context Protocol (MCP)](https://github.com/modelcontextprotocol) - Integration standards353- [AI Agent Best Practices](https://github.com/ai-safety-standards) - Safety guidelines354355### Tools & Frameworks356- [Grok Agent Framework](https://grok.com) - Primary target platform357- [Mistral Vibe](https://github.com/mistralai) - AI agent CLI358- [Agent Skills Ecosystem](https://agentskills.io) - Skill discovery and sharing359360### Documentation361- [Agent Skills Documentation](https://agentskills.io/docs) - Official documentation362- [MCP Documentation](https://github.com/modelcontextprotocol/specification) - MCP standards363364---365366## 🎯 Key Differentiators367368| Feature | This Repository | Other Skill Repositories |369|---------|----------------|--------------------------|370| **Safety Focus** | ✅ Core principle - all skills have safety checks | ❌ Varies by contributor |371| **Consistent Structure** | ✅ All skills follow same format | ❌ Inconsistent formats |372| **Comprehensive Testing** | ✅ Testing guide and requirements | ❌ Minimal testing |373| **Publication Safeguards** | ✅ 6 mandatory controls | ❌ No controls |374| **Auto-Generated Catalog** | ✅ SKILLS_INDEX.md from live metadata | ❌ Manual catalog |375| **Validation Pipeline** | ✅ Automated checks | ❌ No validation |376| **Documentation Quality** | ✅ 10/10 standard | ❌ Varies widely |377378---379380## 📌 Important Notes381382### Design Principle383> **Skills should make agent behavior more useful, auditable, and safe.** They are guidance packages, not permission to bypass controls, disclose private data, or automate high-impact actions without human approval.384385### Usage Warning386> **Always read and understand a skill before using it.** Check the triggers, boundaries, and warnings in each SKILL.md file. Follow all security requirements in [SECURITY.md](./SECURITY.md).387388### Contribution Warning389> **Do not submit skills that**: bypass authentication, evade CAPTCHAs, circumvent paywalls, harvest credentials, automate spam, violate platform terms, or claim unsafe capabilities.390391---392393## 🏆 Version History394395| Date | Version | Changes |396|------|---------|---------|397| 2026-09-11 | 2.0.0 | Added comprehensive documentation (SKILL.md, TESTING.md, STATUS.md, CONTRIBUTORS.md), pre-commit hooks, markdown lint config, secrets baseline |398| 2026-09-10 | 1.2.0 | Enhanced documentation (SECURITY.md, CONTRIBUTING.md, CODE_OF_CONDUCT.md, CHANGELOG.md, GOVERNANCE.md, PUBLISHING.md, SUPPORT.md), added .github workflows and templates |399| 2026-07-30 | 1.0.0 | Initial repository with 89 skill packages, catalog system, validation pipeline |400401---402403*Last updated: September 11, 2026*404*Maintainer: Stijnman*405*Repository: [grok-custom-skills](https://github.com/Stijnman/grok-custom-skills)*406407---408409**Need help?** See [SUPPORT.md](./SUPPORT.md) or open a GitHub issue.