AI & ML

to-aoki Bundle

Use this skill when the user wants to browse a directory that contains many Agent Skills, choose a subset that matches a request, and install the selected skill folders into workspace directories such as `.github/skills` or `.claude/skills`. Use it for cloned skill collections such as `awesome-copilot/skills` and for copying up to a requested count of skills at a time.

to-aoki Bundle

Use this skill when the user wants to turn one or more Markdown design, requirements, architecture, API, schema, ADR, or implementation-plan files into a new Agent Skill for a software development company. The generated Skill must help an agent read Markdown source material and then write or modify production code using a reference-backed implementation playbook, especially when the result should separate operational guidance in SKILL.md from detailed engineering references.

harudium Skill

Use when attachment OCR was added to agent_parser and you need to validate or finish the rollout end-to-end, especially message-turn duplication, conversation-context identity drift, Paddle vs model-api OCR behavior, and commit readiness.

harudium Skill

Use when Threat Stream rows, sidebars, summaries, or live feeds disagree after re-scan, especially for prescan-blocked rows that should transition to a new final state.

harudium Skill

Use when model-api is deployed but scan traffic still fails, especially ALB/TG health, EFS path issues, vLLM sidecar startup, `/v1/scan` 500s, or legacy model-selection 404s.

harudium Skill

Use when conversation summary generation is slow or keeps polling for short threads, especially when the system already knows there are fewer than the minimum required user turns before summary generation should start.

ray-suton Bundle

Run structured multi-agent discussion, critique, and synthesis for product, engineering, research, and operations decisions. Use when the user wants debate, red teaming, sanity testing, launch-readiness review, workflow critique, architecture tradeoff analysis, or a sharper recommendation than a single-perspective answer would provide.

stoffl6781 Skill

Generiert Pest-Tests für Laravel. Scannt Codebase, erstellt Tests nach Priorität. Nutze bei: "Tests schreiben", "Test generieren", "Pest Tests", "Feature Tests", "Unit Tests", "teste den Controller/Model/Service", "Coverage verbessern", "ist das getestet?", "Refactoring absichern".

travisleeeeee Bundle

Core Mission

travisleeeeee Bundle

🎯 Your Core Mission

travisleeeeee Bundle

Your Core Mission

travisleeeeee Bundle

🎯 Your Core Mission

aymenfurter Skill

Create an ad-hoc Microsoft Foundry agent for a specific purpose using the new v2 Responses API. Build a custom prompt agent with optional tools, run a conversation, and clean up. Use when the user says: "create a foundry agent", "make an agent for", "foundry agent", "build an ai agent", "create an assistant", "spin up an agent", "agent for this task"

esurovtsev Skill

Use this skill for any GitHub-related task. Always read it before using gh. It contains the default commands and operating rules for working with GitHub through the local gh CLI.

esurovtsev Bundle

Use when you need to work with GitHub through Python MCP wrappers. The agent should discover the needed wrapper files, run Python inline with the skill's venv Python, and return only compact results.

zeon-kun Skill

Prompt the user to save the current agent output as a markdown spec file. Handles y/n prompt, target directory selection, filename generation, and file writing. Use at the end of any agent workflow.

esurovtsev Skill

Guidelines for systematically exploring and understanding directory structures.

esurovtsev Skill

How to find, read, and audit configuration files — includes concrete investigation steps like grepping for env vars, checking for hardcoded secrets, and mapping external service dependencies.

prasad-vennam Skill

Implement local Retrieval-Augmented Generation using Room and Vector embeddings.

zhouhexin Bundle

小红书技能（MCP版本）

asv-digital Skill

Redija prompts que transformam IA em designer pessoal. Midjourney, DALL-E, Stable Diffusion, Flux. Estrutura, estilos, aspect ratios, negative prompts.

kriszi-soon Bundle

Security auditor for Claude Code skills, plugins, and SKILL.md files. Scans any skill file, skill link, or pasted skill content for malicious code, prompt injection attacks, data exfiltration attempts, dangerous shell commands, hidden Unicode tricks, and social engineering patterns that could harm your system or compromise Claude's behavior. TRIGGER when user says "scan this skill", "check this file for malicious code", "is this skill safe", "audit this plugin", "check for prompt injection", "verify this before I install it", pastes a GitHub link to a skill, or shares any .skill / SKILL.md / plugin content and wants a safety check before installing it.

233i Skill

发现并调用 agent skill。适用于会话开始时，或当你需要判断当前任务该使用哪个 skill 时。这是管理所有其他 skill 如何被发现与调用的元技能。

233i Skill

优化 agent 的上下文配置。适用于开始新会话、agent 输出质量下降、在不同任务间切换，或你需要为项目配置规则文件与上下文时。

233i Skill

记录决策与文档。适用于做架构决策、修改公开 API、发布功能，或需要为未来的工程师与 agent 留下理解代码库所需背景时。

233i Skill

执行多维度代码评审。适用于合并任何改动之前，也适用于评审自己、其他 agent 或人类写出的代码。当你需要在改动进入主分支前从多个维度评估其质量时，就用这个 skill。

kim-dh2417 Bundle

Use when the user wants clear sub-agent role separation for project work, especially around planning, design, development, and QA responsibilities. This skill defines what each role owns, which built-in agent type to use for each role, how to split work cleanly, and how to avoid overlap between parallel agents.

233i Skill

在真实浏览器中进行测试。适用于构建或调试任何运行在浏览器中的内容。也适用于需要检查 DOM、捕获控制台错误、分析网络请求、做性能剖析，或通过 Chrome DevTools MCP 用真实运行时数据验证视觉输出的时候。

jonatangs777 Bundle

Uses vector and matrix reasoning for transformations, constraints, and model formulation.

jonatangs777 Bundle

Applies statistical inference, estimation, and hypothesis testing with explicit assumptions.

jonatangs777 Bundle

Reasons about concurrency using ordering, visibility, and memory-model constraints.

madeinlowcode Skill

Persistent knowledge base for Claude Code projects using the LLM Wiki methodology by Andrej Karpathy. Bootstraps a structured wiki (markdown files with cross-references) and configures Claude Code to maintain it automatically across sessions. Integrates with Obsidian via MCP for graph view and visual navigation. Use this skill whenever the user wants to: set up a project wiki, create persistent memory for a project, use the Karpathy wiki method, organize knowledge with Obsidian integration, bootstrap a knowledge base, or mentions "wiki", "knowledge base", "persistent memory", "LLM wiki", or "Karpathy wiki". Also trigger when the user says "wiki ingest", "wiki query", "wiki lint", "wiki status", "wiki decision", "wiki learning", "wiki search", or "wiki sync".

madeinlowcode Bundle

This skill should be used when developing complex projects that span multiple sessions or context windows. It implements the Long-Running Agent Protocol based on Anthropic's recommended practices for maintaining progress across sessions using SubAgents with isolated context windows, progress tracking files, and incremental feature development.

madeinlowcode Bundle

Assists in creating optimized, well-structured prompts for the ralph-loop plugin. Use this skill when (1) about to use /ralph-loop and need to craft an effective prompt, (2) need to validate a prompt for ralph-loop compatibility, (3) want templates for common ralph-loop use cases (TDD, feature dev, bug fixing, greenfield projects), or (4) need to ensure prompts have clear completion criteria and proper <promise> tags. Ensures prompts follow ralph-loop best practices with measurable success criteria and prevents errors that could compromise project execution.

madeinlowcode Bundle

This skill should be used when the user asks to "create a prompt for Lisa", "generate PROMPT.md", "write specs for Lisa loop", "prepare Lisa execution", mentions "Lisa plugin", "iterative loops", "spec-based verification", or discusses creating high-quality prompts with completion promises and verification steps for the Lisa Claude Code plugin.

madeinlowcode Bundle

Generates comprehensive sprint documentation (spec.md, plan.md, tasks.md, research.md, features.xml) following Spec-Driven Development methodology with multi-persona analysis (Architect, Dev, QA, Designer, PM, BA). Use when planning new features or refactoring existing ones. Compatible with long-running-agent. Automatically researches documentation (WebSearch/WebFetch), creates 50-100 granular tasks with E2E (Playwright) and TDD coverage, generates sequential feature IDs (FEAT-XXX), and creates git branch/commit. Always asks interactive questions to gather requirements before generating complete, production-ready context.