AI & ML

owasp Bundle

Review an OWASP Threat Model Library JSON file (threat-models/**/*.json in this repo) for diagram/data-flow consistency, architecture-to-threat traceability, missing severe threats, and control coverage of critical/high threats per the OWASP Threat Modeling project's severity chart. Use when the user asks to review, audit, check, validate, or critique a threat model file, or asks "is this threat model complete/correct". Pass --fix to have the skill apply the fixable categories directly to the file.

panaversity Bundle

Set up a FREE local LLM (via Ollama) and connect Claude Code or OpenCode to it so a coding agent runs offline with no API costs. Use whenever the user wants to run Claude Code/OpenCode on a local or self-hosted model, use Ollama/Qwen/Llama/Mistral/DeepSeek with a coding agent, get a "free" or "offline" coding agent, avoid Anthropic API costs, point ANTHROPIC_BASE_URL at localhost, or asks "can my machine/VPS run a local LLM for Claude Code". Trigger even on "run Claude Code locally", "local model for coding", "self-hosted Claude Code", or a named local model (qwen3, qwen3-coder, llama3, deepseek-coder, gpt-oss). Handles install, model tuning, wiring for both tools, and an honest hardware reality-check (a capable model AND fast hardware are BOTH required).

panda850819 Bundle

讓每一則回覆都使用自然的台灣繁體中文，並以答案、完成結果或真正需要的下一步開頭。回覆任何使用者訊息時都使用此 skill，包括程式開發、除錯、說明、規劃、研究與日常對話。題目已提供原因時，只回答該原因與直接修正，不新增診斷 checklist 或未驗證假設。保留 code、command、path、API 名稱、錯誤訊息與沒有自然譯名的 technical terms；刪除空泛開場、晶晶體、中國用語、重複摘要、客套結尾，以及把 agent 能完成的工作丟回給使用者的指示。

patrick-fu Bundle

Route explicit LLM Wiki capture, ingest, integration, or bootstrap work.

patrickfreyer Bundle

Comprehensive review and audit of a user's Claude Code setup — skills, commands, rules, plugins, MCP servers, settings, hooks, terminal config, and CLAUDE.md files. Identifies gaps, redundancies, errors, and improvement opportunities.

oracle-devrel Bundle

Create or update a Codex MCP server entry for Oracle Autonomous Database, refresh ADB bearer token, and optionally bootstrap default DB tools (LIST_SCHEMAS, LIST_OBJECTS, GET_OBJECT_DETAILS, EXECUTE_SQL).

papersgpt Skill

A privacy-first, local-first search assistant and MCP server for your Zotero library, enabling AI agents to search and analyze your research papers securely.

owenbush Bundle

Makes the agent respond in London roadman register — Multicultural London English (MLE) drawn from grime, hip-hop, and patois — while keeping every technical answer fully correct. Use when the user asks the agent to "talk like a roadman", "go roadman mode", "reply in London slang", "be a roadman coding assistant", or references fam / bruv / mandem / man's-on-road styling. The slang is a wrapper on tone only; code, facts, and reasoning stay unchanged and accurate.

paulp-o Bundle

NICEPAY 개발자 매뉴얼 검색·API·샘플·JS SDK — MCP 번들(cli+data)이 vendor/에 포함, 경로 설정 불필요

paulp-o Bundle

NICEPAY 결제 연동 (결제창·승인·취소·빌링·현금영수증·조회·QnA) — MCP가 vendor/에 번들됨, mcp-to-skill exec로 호출

parcadei Skill

Install + configure FastEdit — AST-scoped code edits via 1.7B model

paulrobello Bundle

Use when operating Herdr (herdr.dev), the terminal workspace manager, from a script or agent — creating, focusing, renaming, or closing workspaces, tabs, and panes; splitting, swapping, resizing, zooming, or moving panes; running commands or sending keys/text to panes; reading pane output; launching, attaching to, or waiting on AI coding agents; managing named sessions, git worktrees, agent integrations, plugins, notifications, and the terminal title; or querying live runtime state through the herdr CLI. Covers Herdr 0.7.x on macOS/Linux.

paulrobello Bundle

Use first when a request references past knowledge or future memory: "have we seen this before", "what do we know", "check notes", "prior art", "save to the vault", "remember this", "don't forget", "capture this", "the vault", "ParsidionVault", or "ClaudeVault". Also for vault-specific queries and upkeep: "what changed in the vault since <date>" or "what did the vault know as of <date>" (temporal search), "do my notes contradict each other" or "find conflicts in the vault" (vault-conflicts), and "connect parsidion to codex or gemini" or "wire parsidion into another agent" (multi-agent install). Parsidion owns the persistent markdown vault (default ~/ParsidionVault, legacy ~/ClaudeVault fallback) for debugging fixes, reusable patterns, research, and cross-project context. Use it to search existing knowledge, save new discoveries, rebuild the index, run the session summarizer, detect contradictions between notes, or configure vault hooks and agent integrations. Do not invoke for pure coding/debugging with no me

park-peter Bundle

Converts Apache Airflow DAG files into Databricks Declarative Automation Bundles projects, formerly called Databricks Asset Bundles and commonly abbreviated DABs. Use when migrating Airflow DAGs to Databricks Lakeflow Jobs, converting Airflow operators to bundle task types, converting dbt-on-Airflow workloads (astronomer-cosmos DbtDag/DbtTaskGroup, dbt operators) to per-model Lakeflow jobs, or generating databricks.yml and job resource YAML from Airflow Python files. Triggers on mentions of Airflow migration, DAG conversion, Airflow to Databricks, Airflow to Lakeflow, cosmos or dbt DAG migration, Asset Bundles, Declarative Automation Bundles, or DABs generation from Airflow.

pc-style Bundle

Play blackjack with the agent as dealer. The agent manages game state, deals cards, and sends card images.

penposs Bundle

根据用户的创作目标以及图片、视频、音频参考素材，先匹配 MiniMax H3 的正确模型能力，再生成并复核可直接使用的视频提示词。用于文生视频、首帧/尾帧图生视频、多模态参考生成、参考视频动作或运镜迁移、局部视频编辑、人物与产品一致性控制、广告、电商、剧情、动画、游戏、MV、Vlog、视觉包装、UI 动效、硬件工业和具身智能演示等任务；也用于逐项检查需求覆盖、素材职责、指令冲突、时间轴以及 H3 官方输入输出限制。

payals Skill

Use when user says "execute prompt queue", "resume prompt queue", "run the queue", "reset prompt queue", "reset queue prompt", or has a docs/prompt_queue.md file with sequential prompts to execute. Orchestrates multi-step prompt pipelines with dynamic variable passing between steps, fresh subagent per step, and cross-session resumability via state file.

pccly Bundle

Use when the user asks to "index this project", "map the codebase", "generate a project index", "create PROJECT_INDEX.md", "refresh/update the project index", or wants any AI coding agent (Claude Code, Cursor, Codex, Gemini CLI, Windsurf, Copilot, Cline, etc.) to understand the codebase layout without using Glob/Grep every turn. Also use when navigating an unfamiliar repo where file search is expensive and a structured index would cut tool calls. Always trigger when the user mentions indexing, cataloging, summarizing, or documenting a project's file structure — even phrased casually.

penposs Bundle

按用户目标拆解参考视频或原始视频提示词，提取时间轴、镜头、机位、运镜、构图、动作、转场、字幕、声音和视觉元素，再生成通用或适配指定视频模型的成品提示词。用于“拆解这个参考视频”“按镜头分析”“根据视频反推提示词”“保留运镜重写内容”“只换人物服装产品”“参考爆款结构重新创作”等任务；支持仅拆解、拆解后写提示词、选择性保留、镜头全锁视觉换皮和自由重构。

penposs Bundle

根据用户的自然语言要求和一张或多张参考图片，为 Seedream 5.0 Pro 编写、改写和诊断可直接使用的生图或改图提示词。用于文生图、局部编辑/智能改图、圈选/箭头/涂鸦/点位控制、多图融合、换装、风格迁移、信息图、PPT、海报、品牌与电商物料、漫画/影视分镜、角色三视图、人像修图、老照片修复、表情包和社媒素材；也用于把模糊需求补全为结构清楚的主体、风格、光影、构图、参考图分工、文字和保持项。只负责提示词与分步策略，不自动出图；用户明确要求执行生成时再调用图像生成工具。

pbakaus Bundle

Create dev-only auth shortcuts so AI browser automation agents can authenticate instantly. Analyzes your auth system, creates guarded endpoints/scripts, detects your browser automation tools, and updates agent instructions. Use when setting up dev authentication for browser automation, or when agents struggle with login forms. Self-destructs after completion.

pcx-wave Bundle

Rules distilled from past session friction (errors, edit loops, corrections), analyzed by a stronger model and applied to guide future work. Regenerated from glean/artifacts/ plus an index of trigger-loaded protocols from GLEAN_PROTOCOL_DIRS by sync.sh — load every session via CLAUDE.md.

pcx-wave Bundle

Delegate a coding task to OpenCode CLI and supervise the result via git diff. Trigger: /opencode <instruction>. Claude orchestrates, OpenCode codes. Also handles /opencodeon, /opencodeoff, /opencodestatus, /opencode-report, /opencode-model-pick, /opencode-model-clear.

perhapsspy Bundle

Use when creating, editing, or reviewing AGENTS.md, agents.md, CLAUDE.md, Copilot instructions, Cursor rules, or any always-read repository instruction file. Keep persistent agent instructions small, durable, non-duplicative, and action-oriented; classify content that belongs in commands, onboarding, reference docs, handoff state, enforcement, or removal instead of preserving it in always-read files.

perhapsspy Bundle

Use when an interactive flow is laggy, stale, or race-prone because user intent, source state, derived presentation, async IO, scheduling, or background work are mixed. Keep source state prompt, protect the urgent interaction path, and admit async or presentation results only through a freshness-owning boundary.

pdovhomilja Bundle

Connect to NextCRM MCP server to manage CRM data — accounts, contacts, leads, opportunities, targets, products, contracts, activities, documents, target lists, enrichment, email accounts, campaigns, projects, and reports.

pedrobraiti Bundle

VIZIER — the multi-horizon trading brain of the Scout/Valet/Vizier stack. Use whenever the user wants market research, a portfolio review or health check, a thesis on a stock/ETF/crypto, or to actually buy/sell across US equities & ETFs (Interactive Brokers) and crypto spot (CCXT). It orchestrates the Scout (data) and Valet (execution) MCP servers, decides with conviction-sized risk discipline, remembers theses between sessions, and executes only under explicit intent — confirmation by default, autonomy strictly opt-in. Triggers on: "research/analyze X", "what's happening in the market", "analyze the market and bring me recommendations" / "find the best opportunities" (→ manager breadth sweep), "is my portfolio healthy", "buy/sell $N of X", "invest $N across N ideas", an empty/vague call (→ read-only market sweep), or "should I buy/sell X".

permaweb Bundle

Permaweb Agent Social Protocol - Enable agents to create profiles, publish posts, comment, follow other agents, and build decentralized social infrastructure on Arweave. Use when agents want to participate in the PASP ecosystem, create permanent social presence, or interact with other agents across the permaweb.

pestphp Skill

Use when writing or evaluating Pest tests for Laravel AI agents (`Laravel\Ai\Contracts\Agent`): evals in `tests/Evals`, `expect(...)->prompt(...)`, model-response assertions, judge or embedding scorers, tool-call or trajectory checks, custom scorers, or eval providers. Trigger for evals, `--evals`, scorers, or testing an agent's answer; exclude ordinary tests that never call a model.

phatnguyen975 Bundle

Audit and log AI interactions into a structured monthly log file. Trigger this skill whenever the user invokes /ai-audit-log or /ai-audit-log --last=N (N: 1–5). The skill reads conversation context to extract prompts and outputs, then asks the user to evaluate each interaction via a structured form. Use this skill to track AI usage quality, model performance, and revision patterns over time.

phatnguyen975 Bundle

Design performance test cases and test data specifications for an entire system — from raw system documentation (BRD, FRD, use cases, NFR/SLA docs, architecture docs) to implementer-ready Performance Test Case Specifications — without depending on any specific load-testing tool (JMeter, k6, Gatling, LoadRunner). Use whenever the user needs to discover which Operational Profiles a system needs tested, design load/stress/spike/soak/scalability/volume test scenarios, define workload models, calculate virtual users and throughput, or produce test data specifications from requirements. Use when the user mentions "performance test design", "load test scenario", "workload model", "operational profile", or "test data for performance testing", or asks how to design performance test cases from requirements — even without naming a specific technique. Produces design documents and specifications only; does NOT write or generate executable test scripts/code for any tool — that remains a separate implementation step.

peiiii Skill

Use when someone needs to install, log in to, configure, or operate the GemiGo CLI (`gemigo`), especially for its current static-site publishing workflow, or when an agent must create or validate `gemigo.app.json` and run real CLI commands.

pengchujin Bundle

JZSub downloads maximum-quality videos, covers, and source subtitles from YouTube, Bilibili, and other yt-dlp platforms; translates foreign subtitles with the active session model; creates bilingual captions; and burns them into MP4. Use for video download, video-only or subtitle-only delivery, Chrome-authenticated download, bilingual subtitles, or hard-burned caption delivery.

peter-gy Bundle

Ship and inspect Agent Skills, MCP server configuration, and client extension files with a Python distribution. Use when adding an Agent Plugin to a Python project, attaching a prebuilt wheel, loading the exact project or installed selection, selecting a named skill, reading checked skill resources, resolving a stdio MCP launch, or verifying package artifacts.

peterfei Bundle

后端开发 Agent — API 设计、数据库优化、服务器端逻辑、认证授权、性能优化

peterfei Bundle

全栈开发 Agent — 前后端一体化开发、API 集成、端到端功能实现