Recently added

Refactor code using the Refactoring Guru / Fowler catalog: detect code smells, apply the matching named refactoring in small behavior-preserving steps, verify after each.

Enterprise autonomous security auditing, zero-false-positive PoC verification, and context-aware patch remediation using Google Cloud CodeMender (cm) on Gemini Enterprise Agent Platform.

User-global Claude Code hooks, loaded in place as global-hooks@skills-dir through the ~/.claude/skills symlink

Probe plugin for the R3 function-hook question: answers tool.list with one named tool removed and writes one marker file per call. Used by tests/probes/function-hooks-probe.sh.

Portable, harness-agnostic governance lifecycle for the Ubiquity substrate (this plugin is the Claude Code adapter). Captures lessons as CogPRs, preserves session continuity, routes proposals through human-gated review, and tracks recurring friction through the signal manifold.

Rà soát rủi ro hợp đồng thi công xây dựng và cơ điện tại Việt Nam, đối chiếu với kho tri thức 319 điều khoản đã được chuyên gia thẩm định trên 7 dự án thật (FIDIC và mẫu hợp đồng Việt Nam).

PyTorch's API in a browser tab — pick the right borch entry point, follow the ten rules, and start from a recipe that the site's own checks run every night.

University presentation research, planning, PPTX generation, and review suite with shared standards, image strategy, and anti-AI wording checks.

Context operating system for AI coding agents: the agent states intent and the OS orchestrates AST slicing, out-of-context commands, derived architecture indexing and evidence capture.

Expert workflows for the Roark MCP: register a voice agent, design personas and flows, build simulation run plans, and configure the metrics that grade them.

Lint the things a linter cannot: names that drifted from their implementation, comments that became false, tests that do not verify what they claim. ast-grep picks the code, one sentence is the rule, a model answers it. Includes the shipped rule packs, a cookbook for writing your own, and the calibration procedure.

Open-source MCP server for AI agents: web search, content extraction, and library docs.

提示词工程技能包：交付 Loop/Graph 设计器（loop-graph-designer）+ 项目宪法生成器（constitution-generator）+ 会话交接器（session-handoff）。

Structurizr DSL diagram composition, C4 model visualization, and architecture documentation plugin

Provider-backed workflow over GitHub Issues, Jira, GitHub repository wiki directory, and Confluence with issue and knowledge authoring contracts

PlantUML diagram reference, creation, and validation plugin

Extract and organize Open API specifications from KOSIS (Korean Statistical Information Service)

Write a self-contained session handover that a later session can resume. Claude Code carries a recorded handover across /clear.

User-configured answer, turn, and plan reviews, glob-based file reviews, and Git diffs in Herdr.

HTML/XML DOM structure analysis toolkit with hierarchy visualization and CSS selector discovery

A plugin for reading and searching various document formats including CHM files

D2 diagram composition, rendering, and visualization plugin

Tooling for the workflows that Abildtoft encounters in his daily work

Low-level agentic design patterns. Turning old engineering wisdom into reflexes your agent reaches for on its own—on any agent.

Tend's internal CI skills — reviews PRs, triages issues, fixes CI, sweeps repos for improvements. Loaded by tend's GitHub Action; not for general manual use.

Set up and diagnose Tend, an autonomous junior maintainer for GitHub repositories.

Verbatim Jev-guided compaction for Claude Code sessions.

Portable coding lifecycle for AI agents: deterministic routing, research, planning, test-first changes, delegation, verification, review, security, release, handoff, evaluation, and recovery across 27 canonical skills.

Agent-driven terrain and hard-surface asset pipeline for a city builder

Domain modeling, call-graph tracing, and test-driven implementation

Evidence-based code-smell reviews and focused refactoring, informed by Refactoring.Guru

Skills for agent personas: role-based working procedures for coding agents.

将课件整理为分层 Mermaid 思维导图和可互链的 Obsidian 概念笔记。

Standalone adaptive workflow, thin Z80 routing, BridgeZX delivery, and specialist development, documentation, porting, debugging, audit, organization, size, and optimization skills.

Командный plugin с переиспользуемыми Codex skills, практическими примерами и registry-метаданными.

Equipo de agentes PM, arquitectura, programación, QA y seguridad accesible por MCP, REST y CLI.

Evidence-first product and SEO research, AI/tech topic discovery, and X drafting for solo operators.

Shared Agent Skills, packaged for Claude Code and Codex from one canonical implementation.

Shared Agent Skills, packaged for Claude Code and Codex from one canonical implementation.

한국농수산대학교 학교 공식 작성지침에 맞춰 창업논문 초안 작성·검토를 돕는 로컬 스킬. 생성된 결과물은 지도교수 승인과 학교 제출 절차를 거쳐야 합니다.

Agentic Harness skills for repository architecture, audits, migrations, model fit, security, design intelligence, delivery, bounded decisions and context efficiency.

A context-aware German and English writing, editing, intake, and lightweight research system.

rust-script 渗透测试体系:五域(资产侦察/漏洞扫描/漏洞利用/后渗透操作/攻击产物分析)单文件脚本、pentest-lib 框架、wssd 远程壳 daemon、vulhub 实证回归。仅限合法授权测试。

Independent whole-repository reviews, verified fixes, checks, and authorized local commits.

Project-scoped todos, scratchpads & plans for you and your agents, plus the tally dev loop: brainstorm → plan → build → review-branch, with debug and setup.

Elliott's personal AI-founder OS: department agents (Strategy, Product, Marketing, Sales), venture workflows, and a portfolio hub. Free, local, built on Claude Code.

Agent skills by JaysonGCS. Includes 1 skill.

Phase-gated workflow enforcer, cross-project experience memory, and git snapshots for Claude Code

Voltra trainer control as an MCP server plus a push-event channel

Desenhos SVG (escolar, técnico, lúdico) com loop de auto-correção visual via Inkscape MCP.