⏳ This skill is pending AI review.

Scores will appear once the review pipeline completes.

v3.2.0

hardshell

@asistent-alex⭐ 1 stars

Apply when writing, reviewing, refactoring, or designing any code.

Use with your AI agent

Open your project in any AI assistant that can read your files. Works with ChatGPT, Claude, Claude Code, Codex, Cursor, Hermes Agent, OpenClaw, Grok Bot, and more.

Your agent needs access to this page’s linked instructions and your project files. Copying does not install or execute anything.

—/10

// RATINGS

⭐GitHub Stars
⭐ 1 on GitHubGitHub ↗

New / niche

🟢ProSkills Score
—
📍

Not yet listed on ClawHub or SkillsMP

// README

Hardshell - OpenClaw Coding Standards

Version: 3.2.0

A comprehensive coding standards skill for OpenClaw agents. Apply every rule on every coding task, without exception.

Installation

# Install from ClawHub (when published)
openclaw skill install hardshell

# Or install from GitHub
openclaw skill install https://github.com/asistent-alex/openclaw-hardshell

What It Covers

Security (Highest Priority)

  • Input validation and trust boundaries
  • Secrets and credentials management
  • Authentication and authorization
  • Common vulnerabilities (SQLi, XSS, CSRF, SSRF, path traversal)
  • API and HTTP security
  • Dependency auditing

Architecture & Design

  • SOLID principles
  • Separation of concerns
  • Law of Demeter
  • API design patterns
  • Error handling strategies

Code Quality

  • Naming conventions
  • Function design
  • Comments and documentation
  • Code smells to avoid

Testing

  • TDD (Test-Driven Development)
  • Test pyramid (70% unit, 20% integration, 10% E2E)
  • Coverage minimums (70% on business logic)
  • Test isolation rules

Git Workflow

  • Branch strategy (trunk-based with short-lived feature branches)
  • Conventional commits format
  • PR process and code review

Performance

  • Complexity analysis
  • Caching strategies
  • Optimization guidelines

Agent Behavior

  • Think before coding — surface assumptions and tradeoffs
  • Simplicity first — minimum code, nothing speculative
  • Surgical changes — touch only what you must
  • Goal-driven execution — define verifiable success criteria

Usage

When the agent is working on code, it automatically loads relevant reference files:

User: "Write a Python function to send email"
Agent: Loads references/languages/python.md for Python-specific standards

When editing existing code or receiving an ambiguous request:

Agent: Loads references/agent-behavior.md for surgical change rules

Reference Files

FilePurpose
references/agent-behavior.mdBehavioral guidelines for LLM coding agents
references/git-workflow.mdBranches, commits, PRs
references/performance.mdComplexity, caching
references/skill-development.mdStandards for creating OpenClaw skills
references/testing.mdTDD, test pyramid
references/languages/python.mdPython-specific
references/languages/typescript.mdTS/JS-specific
references/languages/go.mdGo-specific

Validation

Run the smoke test after any change:

bash scripts/validate.sh

Changelog

See CHANGELOG.md for version history.

License

MIT License - see LICENSE

Contributing

This skill is designed to be universal. If you have improvements that apply across all OpenClaw skills:

  1. Fork the repository
  2. Create a feature branch
  3. Submit a pull request with your improvements

Author

OpenClaw Community

Acknowledgments

The agent behavior guidelines in references/agent-behavior.md are derived from Andrej Karpathy's observations on LLM coding pitfalls, implemented via the community skill multica-ai/andrej-karpathy-skills.

Links

// HOW IT'S BUILT

KEY FILES

SKILL.mdREADME.md

// REPO STATS

1 stars