SoonYub Hwang
Community@nowonbun
Welcome to my github. This github is not completed for the several source. Thank you.
Agent Skills by SoonYub Hwang
Showing 21 vetted skills indexed across 1 GitHub repositories.
stock-analysis
Synthesize company fundamentals, technical indicators, and market context into a risk-aware investment conclusion.
markdown-safe-writing
Preserve UTF-8 Korean text when editing Markdown via Windows PowerShell.
skill-governance-rule
Classify skill governance risks and assign binary control tiers for sensitive operations.
skill-create-rule
Validate SKILL.md files against frontmatter, header, and section rules.
engineer
Prioritize root-cause hypotheses with test plans and rollback criteria.
ai-agent
Define AI agent operational requirements and decision boundaries for deployment.
skill-modify-history
Record causes, actions, and preventive measures into standardized history artifacts.
stock-mcp
Validate market and date parameters and select the correct StockSearcher MCP tool.
reality-check
Identify missing assumptions and downside risks in optimistic proposals.
mariadb-mcp
Validate MariaDB MCP connection scope and read/write permissions before executing SQL.
work-runtime
Enforce runtime controls for delegated agent workflows with validation and reporting.
automation
Design and evaluate automation workflows for scripts, CI/CD, and scheduled tasks.
skill-check-rule
Audit skill documents for consistency, references, and governance alignment.
claude-review-runtime
Preflight MCP availability and delegate code, documentation, and configuration changes to the claude-cross-review-protocol.
coding-assistant
Classifies code and config changes as destructive/non-destructive and coordinates Claude review sequences with verification reporting.
vocabulary-rule
Normalize ambiguous vocabulary and rule-strength terms in governance documents.
github-mcp
Validates GitHub MCP targets and separates remote from local git operations.
research
Analyze topics by defining questions, separating facts from interpretations, and reporting uncertainty.
ai-collaboration-governance
Enforce runtime controls for Claude MCP collaboration with timeouts and healthchecks.
claude-cross-review-protocol
Automate Claude cross-review workflows with MCP-based governance and standardized logging.
constitution-rule
Define governance rules for global_instructions and AGENTS with priority ordering and stop conditions.