lauvsongCommunity·9 Agent Skills Included

claude-soul

Safety-first coding rules, guardrails, and review workflows

Enforces strict safety rules, approval protocols, and verification steps for everyday coding work. Blocks destructive commands, secret leaks, and test bypasses through hooks and permission policies. Includes skills for codebase Q&A, GitHub CLI operations, design stress-testing, Korean tone review, and session handoffs. Optimized for Kotlin and WebFlux backend development environments.

35|3

↗

npx skills add lauvsong/claude-soul --all -g -y

Available:

Directs the agent to respond in Korean, follow strict safety and approval rules, classify task risk before acting, and load detailed rule files for workflow, coding discipline, and stack-specific conventions.

All Skills in This Repository (9)

Pure Emerald Level Indicators

📦 In Repo

lauvsong

github-notification-cleanup

Filter GitHub notifications to keep open reviewer Pull Requests and archive others.

brew-upgrade

Automate upgrading all Homebrew packages on macOS with brew upgrade.

gh

Automate GitHub.com tasks using the gh CLI for PRs, issues, reviews, and releases.

ask

Analyze codebase structure, behavior, and dependencies with source-backed explanations.

codex

Automate non-interactive code analysis and transformation using Codex CLI.

template-skill

Generate a YAML frontmatter and Markdown body skeleton for new Claude skills.

claude-security-review

Audit Claude Code settings.json, guardrail.py, and safety rules for security gaps.

lesson

Capture user feedback as standardized, deduplicated lessons appended to tasks/lessons.md or echoed in chat when needed.

lesson-global

Apply user feedback to global rules files under ~/.claude/rules.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install claude-soul?▼

Run `npx skills add lauvsong/claude-soul --all -g -y` in your terminal to install all skills in this suite globally.

What does claude-soul do?▼

It applies a safety-first rule set to your coding agent, blocking destructive commands and secret exposure while enforcing verification before any task is marked complete.

Which tech stack is claude-soul designed for?▼

It is tuned for Kotlin and WebFlux backend development, with dedicated rules for reactive programming, testing, and coding discipline. Other stacks work but may need rule adjustments.

How does claude-soul prevent dangerous commands?▼

A guardrail plugin and policy file block risky operations like git push, secret file access, and infrastructure changes, requiring explicit user approval before execution.

Does claude-soul support Korean language workflows?▼

Yes. It enforces Korean responses and includes a tone review skill that removes machine-translation patterns from Korean documents and messages.