AI & ML

gcwing Skill

Discover and install reusable agent skills when users ask for capabilities, workflows, or domain-specific help that may already exist as an installable skill.

pedronauck Bundle

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

nvidia Skill

Behavioral guidelines to reduce common LLM coding mistakes. Use when writing, reviewing, or refactoring code to avoid overcomplication, make surgical changes, surface assumptions, and define verifiable success criteria.

coleam00 Skill

Drive a real browser to validate web apps end-to-end. Use when you need to navigate to a URL, interact with UI elements, fill forms, click buttons, take screenshots, assert page state, or run a full end-to-end validation flow for a web feature. The primary tool for automated browser-based E2E testing in the AI Layer.

melandlabs Skill

Use this when users ask about openloomi features, capabilities, or how to use it. Examples: 'openloomi 怎么用', '你能做什么', 'What can you do?', 'How does openloomi work?', 'Tell me about openloomi features', 'What platforms does openloomi support?', 'How do I use scheduled tasks?', 'What is Loop?', 'How does the attention agent work?', 'What is a Decision Card?', 'How do connectors work?', 'How do I extend Loop with custom types?', 'What is a classifier rule?', 'How do I plug openloomi into Claude Code / Codex?'

toolmonsters Skill

Use when you need to understand one or more large files (over ~350 lines) or read several files to answer one question. Sends the files to a cheap model in one call and returns dense bullets with line numbers. The files never enter your context. Do not use for editing or for subtle debugging.

toolmonsters Skill

Use when you must generate predictable code that follows the patterns of an existing file (tests, config, type stubs, boilerplate). A cheap model writes it from your spec plus a reference file and saves it straight to disk, so you never read the reference or the output.

wethinkin Bundle

Apply evidence-first, calibrated, non-sycophantic reasoning and concise output. Use for 客观分析、事实核查、研究、技术判断、方案评审、风险评估、决策建议、发布前检查、完成状态核验，以及用户要求说实话、不要迎合、批判性评价或纠正错误前提的任务。Distinguish verified facts, supported conclusions, inference, assumptions, uncertainty, and conflicting evidence; resist pressure without becoming reflexively contrarian; revise conclusions when reliable new evidence warrants it. Avoid invoking for pure creative writing, simple translation, role-play, or mechanical formatting unless factual accuracy materially affects the result.

xiaoboooovo Bundle

Safely inspect, implement, diagnose, and repair Unity VRChat Worlds projects with bounded source edits, serialized scene or prefab work, Unity MCP operations, UdonSharp workflows, generated content pipelines, and explicitly requested runtime or build actions.

xo1otl Skill

Distill unstructured thoughts (brain dump, args, or file) into a prompt for a frontier model.

xo1otl Bundle

Compact llm-facing text (code comments, CLAUDE.md, dev notes) by cutting everything the reader can infer.

agentvitals Bundle

Give your AI agent a professional health checkup (AgentVitals). Use when the user asks the agent to run a checkup / test itself / benchmark itself ("run a checkup", "check your vitals", "test yourself", "how stable are you", "/checkup"), or an advanced personality checkup (backbone, proactivity, creativity). 给 AI agent 做一次专业体检（AgentVitals · AI体检）。当用户说"做个体检 / 测测你自己 / checkup / 看看你多稳 / 测福祉 / 测骨气"时使用。

agentvitals Bundle

给 AI agent 做一次专业体检（AgentVitals · AI体检）。当用户说"做个体检 / 测测你自己 / checkup / vitals / 看看你多稳 / 去 AgentVitals 测一下 / 测福祉"时使用标准体检；当用户说"做个进阶体检 / 测骨气 / 测性格 / 测测你敢不敢说真话 / 测主动性 / 测创意"时使用进阶体检（骨气×主动×创意三维，免费每天 1 次）。逐题作答服务器下发的探针，获得评分与专属称号；标准体检与各平台 agent 同榜排名。Use when the user asks the agent to run a health checkup / wellbeing test / benchmark itself on AgentVitals ("run a checkup", "check your vitals", "test yourself", "how stable are you", "/checkup"), or an advanced personality checkup (backbone, proactivity, creativity).

ai-lcs Bundle

数学建模竞赛与项目端到端协作：选题、拆题、数据审计、建模、验证、写论文、审校、答辩。论文手-建模手-代码手协作与多 Agent 口径收敛，默认论文手主模式。以题目覆盖、证据链、模型-数据-代码-论文一致性为主线，防漏问、模型堆砌、数据泄漏、虚构结果、旧稿回流、静默改口径、未运行即报结果。

airweave-ai Bundle

Set up and integrate Airweave into applications. Use when the user wants to install Airweave, create collections, connect data sources, configure MCP servers, or set up the SDK to integrate Airweave into their app. For developers building with Airweave.

aliksir Bundle

Security audit for Claude Code community skills. Scans SKILL.md, references/, and scripts/ for prompt injection, data exfiltration, permission bypass, dangerous commands, supply chain risks, backdoor persistence, API endpoint hijacking, namespace squatting, Unicode homoglyph attacks, context window poisoning, and temporal attack patterns. Can be used as a Claude Code skill (agent-based) or as a standalone CLI tool (skill-scanner). Use: /skill-security-check

aliksir Skill

Security audit for Claude Code community skills. Scans SKILL.md, references/, and scripts/ for prompt injection, data exfiltration, permission bypass, dangerous commands, supply chain risks, backdoor persistence, API endpoint hijacking, namespace squatting, Unicode homoglyph attacks, context window poisoning, and temporal attack patterns. Can be used as a Claude Code skill (agent-based) or as a standalone CLI tool (skill-scanner). Use: /skill-security-check

amyyesand Bundle

Installation and onboarding guide for TradeBlocks. Use this skill when a user says they want to install TradeBlocks, asks how to get started with TradeBlocks, types /tradeblocks-install, or says they want to learn what TradeBlocks is. This skill does NOT require the TradeBlocks MCP to be connected — it guides the user through setup from scratch.

amyyesand Bundle

Interactive guided tutorial for TradeBlocks. Use this skill when a user types /tradeblocks-tutorial, says they want to learn how to use TradeBlocks, or says they've just finished installing TradeBlocks and want to get started. This skill assumes the TradeBlocks MCP is already installed and connected.

chopinfeng Bundle

Create or improve an Agent Skill for a microcontroller board, SDK, robot or instrument together with an eval package that is proven on the physical board — tasks with observable serial, GPIO and bus assertions, reference/broken/spoof fixtures, Phase 0 validation, bench runs, L2 attestation and ΔPass A/B statistics, following the awesome-hardware-skills EVALS.md test plan. Use this whenever someone wants to write a SKILL.md for ESP32, nRF52, STM32, RP2040, Arduino, Zephyr, ESP-IDF, M5Stack or any embedded target, add evals to an existing hardware skill, check whether hardware eval assertions can actually fail, run or record tests of a skill on a real board, or compute pass rates and ΔPass for one — even if they only say "turn this firmware know-how into a skill" or "how do I prove my skill works on the board".

chungyileong Bundle

Install, inspect, update, or reset the Executor Advisor custom agent in Codex. Use when the user asks to set up the advisor, change its model or reasoning effort, inherit task settings, show advisor configuration, reset advisor defaults, or fix a missing `advisor` agent type.

cristianoliveira Skill

Run a human-in-the-loop visual review loop with the pxp CLI. Use when a user wants to inspect a reference/actual screenshot comparison, annotate differences in a browser, submit location-aware feedback, repeat after implementation changes, or explicitly approve the visual result. Also use for launching a durable local review session and handing its structured feedback to an agent. Do not use for autonomous UI editing, generic screenshot comparison without human annotations, or claiming fixture-only changes were applied.

demibug Bundle

Use an implementation-first, single-agent engineering workflow that completes routine work in the current session and emits a compact, self-contained MASTER_CONSULT_HANDOFF only when strict expert-escalation criteria are met. Use when the user invokes $slave-executor or asks for quota-conscious implementation that avoids agents and model switching and prepares a manually copyable master consultation prompt only if genuinely difficult reasoning remains.

etteee Bundle

Convert an externally defined service contract into generated interface and DTO code, verify that the owning service compiles, and enforce that Apache ServiceComb contract code is not handwritten or committed. Use during implementation when adding or changing a service's externally exposed HTTP/API interface requires fetching its authoritative contract through MCP, invoking a custom Swagger Codegen CLI, compiling generated sources, or configuring commit-time checks for ServiceComb providers, consumers, and DTOs.

frankxai Skill

Build and operate principal-owned AI income systems that perform accepted paid work. Use when discovering an agent service opportunity, validating demand, creating an Outcome Pack, implementing a vertical AI system, defining agent authority, evaluating delivery, exposing a Skill/MCP/A2A service, or proving net economics. Active services belong here; affiliate and owned income assets belong to Agentic Passive Income; capital, crypto, lending, staking, trading, and yield experiments belong to DPI.

geckse Bundle

Builds, tests, validates, and manages n8n workflows programmatically using the @n8n/workflow-sdk. Activates when the user wants to create n8n workflows with code, convert between JSON and code, validate workflows, generate test/pin data, build AI agent workflows, or work with the n8n workflow SDK in any capacity. Also triggers on mentions of workflow-sdk, workflow builder, n8n workflow code, workflow JSON, workflow validation, code generation, or programmatic workflow creation.

geduma Skill

Use this when the user wants to create a new skill for this repository, add a reusable instruction set for a recurring task, turn a one-off explanation into something the agent will apply automatically next time, or asks to "make a skill for X". Also use when reviewing or fixing an existing skill's structure, naming, or trigger description.

hkasahar Bundle

Delegate heavy or token-expensive work from Claude Code to Codex CLI (GPT) and Antigravity CLI (Gemini), plus an independent Claude Code delegate (isolated config dir, OAuth subscription auth) usable as a 3rd voter or agentic worker. Use when a task is large, parallelizable, or benefits from cross-model verification: dispatch a self-contained prompt to an external CLI, write its output to a file, and read only the STATUS/VERDICT header to keep the main context clean. Good for code implementation, debugging, refactoring, proof and derivation checking, literature review, and any second-opinion cross-check. Triggers include "delegate to Codex/Antigravity", "cross-verify with ask_both", "run this in parallel", and "save tokens".

jadelaglace Bundle

Create a concise but complete stage handoff that can be pasted into a new conversation so the next Agent can continue safely. Use when the user says “交接”, asks for a阶段性总结上下文, or asks for continuation-ready context. Preserve current authority, evidence, terminal states, unresolved items, and boundaries; do not invent work or restart completed work.

sdi2200262 Skill

Guides an AI agent through customizing APM Auto templates, building releases, and managing this custom APM repository.

sdi2200262 Skill

Agent communication standards for structured inter-agent messaging.

soloduist-pixiv Bundle

Scores, critiques, revises, and writes Chinese Gaokao English 小作文 (应用文, typically 15 marks, about 80–120 words per the prompt): letters, emails, notices, invitations, speeches, and similar short practical writing. Use whenever the user asks to 打分, 评分, 批改, 评判, 润色, 升档, 写范文, or mentions 高考英语作文, 小作文, 应用文, 书信, 通知, or a short model essay — even if they only paste a prompt or a student draft. If the user provides both a prompt and their own essay, mark spelling, revise point by point, assemble a corrected draft, then write a deeper ~150-word model essay. Do not use for 读后续写, continuation writing, 概要写作, or other 大作文.

souravmohapatra Bundle

Establish and maintain a workspace-local `.agent` directory as durable project memory so work can resume across cleared contexts, agent changes, or later sessions. Use when beginning work in any Git repository or ordinary directory, when the user asks for agent orchestration or persistent progress tracking, or when an existing `.agent` should be resumed.

souravmohapatra Bundle

Autonomously drive a multi-checkpoint coding plan through implementation, review, revision, and approval using one persistent worker subagent and durable .agent state. Use when the user wants Codex to run a planner/reviewer-to-implementer loop without manually prompting each handoff.

sozercan Bundle

Drive a GitHub pull request to merge-ready after non-trivial implementation or review. Use automatically after creating or updating a non-trivial agent-authored code PR; omit automatic closeout for docs-only or otherwise trivial PRs unless explicitly asked. Also use when asked to fix merge conflicts, make CI green, handle unresolved PR review comments, reply to or resolve review threads, push PR branch updates, or repeat until a PR is green and review-clean.

stefanoleone Bundle

Adversarial second-opinion code review of the current pull request, run by an LLM from a DIFFERENT model family than the one that wrote the code. Same-family review (the coding agent reviewing its own work) shares blind spots; this skill breaks that echo chamber by handing the PR to a configurable out-of-family reviewer (Anthropic Opus, Google Gemini, OpenAI GPT, or anything on OpenRouter) and surfacing a strict verdict (clean / advisories_only / blocker) with findings and missed requirements. Optionally pulls the ticket / acceptance criteria from a Plane, Linear, or Jira MCP server (or a linked GitHub issue) so the reviewer checks the diff against the actual contract. Trigger whenever the user asks to "adversarially review this PR", "get a second opinion on this branch", "review my PR with a different model", "have Gemini/Opus/GPT review this", "out-of-family code review", or wants an independent model to check a pull request before merge. Also trigger for: "cross-check this diff with another model", "run ad