Instruccions per a IA
Disseny d’instruccions i plantilles per a models d’IA
git-safety-hooks
helderberto
Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations or add git safety guardrails.
save-session
blacktop
Save AI agent session information for later resumption. Use when: (1) Finishing a work session and want to save progress (2) User says "save session", "remember this session", or "save your session ID" (3) Before ending a long coding/debugging session (4) When user wants to be able to resume this conversation later Supports Claude, Codex, Gemini, Copilot, Cursor, and other AI agents.
obsidian
Delphine-L
Integration with Obsidian vault for managing notes, tasks, and knowledge when working with Claude. Supports adding notes, creating tasks, and organizing project documentation.
mcp-server-orchestrator
organvm-iv-taxis
Configure, deploy, and troubleshoot Model Context Protocol (MCP) servers for AI agent workflows. Use when setting up MCP servers, debugging connection issues, managing multi-server configurations, integrating with Claude Desktop/Code/Cowork, or designing custom tool servers. Triggers on MCP configuration, tool server development, Claude integration issues, or agent infrastructure setup.
narratological-algorithms
organvm-iv-taxis
Distill artist/theorist narrative principles into formal, implementable algorithmic frameworks. Use when asked to extract, formalize, or systematize narrative techniques from any storytelling source—filmmakers, writers, theorists, game designers, showrunners. Triggers on requests involving narrative principle extraction, story structure analysis, craft methodology formalization, or creating implementable storytelling protocols.
semantic-compression
MuseLinn
Re-encode verbose prose into a dense telegraphic register — punctuation as connectives, label frames, verbless assertions — without losing normativity or precision. Use when compressing system prompts, tool/function descriptions, skill bodies, or agent instructions; reducing token count or context bloat; making documentation token-efficient for LLM input; or rewriting text in compressed notation.
lsp-manager
bahayonghang
Automatically detect programming languages, configure Language Server Protocol (LSP) servers, and troubleshoot LSP integration issues for Claude Code. Use when users request to set up LSP, configure code intelligence, add language support, troubleshoot LSP errors, or when working with projects that need real-time code diagnostics and navigation.
claude-collaboration
Delphine-L
Best practices for using Claude Code in team environments. Covers skill management, knowledge capture, version control, and collaborative workflows.
tool-prompt-optimization
MuseLinn
Optimize the description prompts an AI agent reads to learn its built-in tools (the .md files under prompts/tools/). Two halves: (1) measure how much of a prompt is already inferable from the tool's JSON parameter schema + name, to prune redundancy with evidence; (2) house authoring rules for what belongs in a tool prompt vs what stays in code. Use when auditing, trimming, writing, or reviewing tool prompts, deciding what schema field descriptions already cover, or testing schema-vs-prompt overlap before deleting prompt lines.
orchestrate
blacktop
Lead-developer workflow for orchestrating sub-tasks with Claude Code CLI (Opus 4.5/Sonnet) using structured handoffs, review loops, and role specialization. Use when Codex should act as lead/architect, plan architecture and file structure, break work into atomic tasks, delegate bulk boilerplate/docs to Claude, and verify/iterate on Claude output.
question-refiner
liangdabiao
将原始研究问题细化为结构化的深度研究任务。通过提问澄清需求，生成符合 OpenAI/Google Deep Research 标准的结构化提示词，完全替代 ChatGPT 的问题细化功能。当用户提出研究问题、需要帮助定义研究范围、或想要生成结构化研究提示词时使用此技能。
continuous-learning-agent
organvm-iv-taxis
Self-improvement patterns for AI agents to learn from feedback, errors, and successful patterns across sessions
handoff
blacktop
Generate optimized handoff prompts for delegating tasks to LLM agents. Use when needing to hand off work to another agent (GPT-5.2/Codex, Claude Opus 4.5/Sonnet 4, or Gemini 3 Pro/Flash), whether for a sub-task/parallel task within the same project or a fresh start handoff to a new agent context. Triggers on requests like "create a handoff prompt", "delegate this task to another agent", "hand this off", or "prepare context for another agent".
mlops-engineer
olehsvyrydov
Senior MLOps Engineer with 8+ years ML systems experience. Use when integrating LLM APIs (Gemini, OpenAI, Groq), building AI pipelines, managing prompts, setting up model serving, implementing AI cost optimization, or building training data pipelines.
claude-skill-management
Delphine-L
Expert guide for managing Claude Code global skills and commands. Use when creating new skills, symlinking to projects, updating existing skills, or organizing the centralized skill repository.
securities-analysis-cn
kingqiu
自动分析A股、港股、ETF基金，生成专业PDF深度分析报告（含AI买卖建议和互联网研究）。 支持单只分析和多只对比：输入证券代码（510300、600519、00700.HK）或名称（贵州茅台、腾讯控股）， 自动识别类型。多只输入时自动生成对比分析报告（含综合评分+说人话解读）。 使用场景：当用户要求"分析某只股票/基金"、"对比XX和YY"、"帮我看看XX的投资价值"时触发。 插件化架构：数据源、AI模型、搜索引擎均可通过 .env 配置切换。
review-skill
corygabrielsen
Review a skill document using specialized reviewers. Each reviewer finds specific issue types.
hr-service-delivery
groeimetai
This skill should be used when the user asks to "HR case", "employee center", "onboarding", "offboarding", "HR service", "lifecycle event", "HR catalog", or any ServiceNow HR Service Delivery development.
chronicle
PrakharMNNIT
"Personal journal intelligence that transforms raw, unorganized thoughts into structured diary entries with psychological analysis. Use when the user provides journal entries, diary text, stream-of-consciousness writing, voice memo transcriptions, or asks to process daily thoughts into a structured format. Produces narrative entries, gratitude extraction, multi-level psychological analysis (surface/medium/clinical), health pattern flags, therapeutic micro-actions, and bridge-to-tomorrow planning. Trigger phrases: 'journal entry', 'diary entry', 'process my thoughts', 'Chronicle', 'daily reflection', 'write up my day'."
document-management
groeimetai
This skill should be used when the user asks to "attachment", "document", "file upload", "document template", "PDF generation", "document workflow", or any ServiceNow Document Management development.
change-management
groeimetai
This skill should be used when the user asks to "create change", "change request", "CAB", "change approval", "change task", "RFC", "normal change", "emergency change", or any ServiceNow Change Management development.
ask-effective-llm-coder
NavanithanS
Best practices for LLM-assisted coding. Declarative workflows, simplicity, tenacity.
llm-eval-designer
CAPHTECH
LLM生成システムの検証設計スキル。assay-kitフレームワークを活用し、LLM特有の失敗モード（幻覚、例への過学習、部分的処理）を考慮した総合的なテストケース設計を支援する。 使用タイミング: - LLMベースのワークフロー/エージェントの評価設計時 - ゴールデンデータセット（golden-dataset.yaml）の設計・拡張時 - 既存テストが特定パターンに過学習していないか検証時 - LLM出力の品質スコアラー設計時 - 「なぜこのテストケースが必要か」の根拠を示す時
topic-collector
zephyrwang6
AI热点采集工具。从Twitter/X、Product Hunt、Reddit、Hacker News、博客等采集AI相关热点内容。当用户说"开始今日选题"、"采集热点"、"看看今天有什么新闻"、"今日AI热点"时触发。聚焦领域：Vibe Coding、Claude Skill、AI知识管理、AI模型更新、AI新产品、海外热点。