skills-by-yigitkonur

43 skills for AI coding agents -- code review, planning, research, browser automation, multi-agent orchestration, framework guides, SDK guides, design extraction, and more.

Install

Install all skills at once:

npx -y skills add -y -g yigitkonur/skills-by-yigitkonur

Install a single skill:

npx -y skills add -y -g yigitkonur/skills-by-yigitkonur/skills/<skill-name>

Skills

Skill	Category	Description
apply-clean-architecture	development	Clean Architecture and DDD in TypeScript
apply-liquid-glass	development	macOS Liquid Glass design system (Tahoe+)
apply-macos-hig	development	macOS HIG design system -- spacing, components, accessibility
ask-review	productivity	Hand off work as a PR with domain-aware self-review, or produce a markdown review doc
build-chrome-extension	development	Chrome extensions with Manifest V3
build-convex-clerk-swiftui	development	Convex + Clerk backends for SwiftUI iOS/macOS apps
build-copilot-sdk-app	development	GitHub Copilot SDK apps in TypeScript
build-langchain-ts-app	development	LangChain.js agents, RAG, and tool-calling
build-mcp-server-sdk-v1	development	MCP servers with @modelcontextprotocol/sdk v1.x
build-mcp-server-sdk-v2	development	MCP servers with @modelcontextprotocol/server v2
build-mcp-use-agent	development	AI agents with mcp-use MCPAgent
build-mcp-use-apps-widgets	development	MCP apps with React widget rendering
build-mcp-use-client	development	MCP clients with mcp-use SDK
build-mcp-use-server	development	MCP servers with mcp-use library
build-raycast-script-command	development	Raycast Script Commands in Python/Bash
build-skills	productivity	Claude skill creation and research methodology
check-completion	productivity	Check what's done vs. claimed done via a 22-status taxonomy; remediate until every in-scope task reaches a terminal status
convert-url-to-nextjs	design	Live URLs or HTML snapshots to Next.js projects
develop-typescript	development	Strict TypeScript patterns and config
do-debug	productivity	Language-agnostic systematic debugging; four phases + Iron Law + 3-fails handoff to do-think Interactive
do-review	productivity	Systematic pull request review
do-think	productivity	Deep-thinking framework — Solo or user-in-the-loop Interactive modes; 4-phase loop with mandatory stress-test trio
enhance-prompt	productivity	Turbocharge LLM prompts with steering, halt conditions, and failure pre-emption
enhance-skill-by-derailment	productivity	Skill quality testing via subagent execution and trace analysis
evaluate-code-review	productivity	Evaluate PR/session/markdown review feedback with verify-before-implement discipline and multi-agent consolidation
extract-saas-design	design	SaaS dashboard visual system extraction
init-agent-config	configuration	AGENTS + REVIEW standardization hierarchies
optimize-agentic-cli	development	Audit or design agent-ready CLIs with iterative feedback loops
optimize-agentic-mcp	development	Audit, optimize, or architect new MCP servers
publish-npm-package	development	npm publishing via GitHub Actions
run-agent-browser	testing	Browser automation with agent-browser CLI
run-codex-exec	orchestration	Parallel codex exec agents in git worktrees with auto-commit + live monitor
run-codex-review	orchestration	Per-branch /codex:review fix loops + /ask-review PR + codex rescue + multi-bot /do-review evaluation
run-github-scout	productivity	Adaptive GitHub repo discovery and shortlisting for concrete needs
run-industry-research	productivity	Industry research corpora with evidence packs and comparisons
run-issue-tree	productivity	Plan and execute GitHub Issue trees with sub-issues and waves
run-playwright	testing	Browser testing with Playwright CLI
run-repo-cleanup	productivity	Sweep dirty tree + unpushed commits + N worktrees into focused private-fork PRs with self-review bodies
run-research	productivity	Technical research with web search, Reddit mining, and multi-agent orchestration
swift-quality-hooks	development	Swift pre-commit hook with SwiftLint + SwiftFormat for Apple platforms
test-by-mcpc-cli	development	MCP server testing with mcpc 0.2.x
test-macos-snapshots	testing	macOS app screenshot validation and drift checks
use-railway	platform	Railway CLI commands, workflows, and version-drift routing

Notes

Each installed skill adds to the context window. Install only what you need.
If you see duplicate skill triggers, that is a known Claude Code issue (#27721).

Contributing

See CONTRIBUTING.md.

License

MIT