Recently Added
Fresh skills just added to the collection. Be the first to try them out!
laifaxin-outreach
by suxuemi
AI-assisted outreach SOP package for foreign trade · Pairs with Laifaxin (来发信) — does NOT operate the 来发信 web UI · Generates customer search queries, benchmark profiling, short hook-driven first-round emails, multi-round sequences, reply-recovery playbooks, and a setup checklist · The user takes the outputs and manually configures them in the Laifaxin web app (variables in output are placeholders — must be replaced via the 【插入变量】 button in the 来发信 editor before sending) · Use when finding overseas buyers, writing cold emails, building sequences, or handling customer replies
claude-code-troubleshooting
by loekraveubetsci-creator
在中国大陆 / 受限网络环境下排查和修复 Claude Code (终端 CLI) 的常见问题，尤其是登录失败、403 "Request not allowed"、OAuth 授权回调失败、代理 (Clash / 小云朵 / ClashX) 配置、模型不可用，以及"每次重开终端就要重新登录"等问题。Whenever the user mentions Claude Code 登录不上 / 终端报 403 / Please run /login / OAuth error / 授权点了没反应 / "克洛德很快就会回来" / 终端走代理 / VPN 节点 / 重复登录 — even casually, use this skill and walk them through the ordered diagnostic steps. 适用于使用 Clash 系代理工具的中国网络环境用户。
brooklet-ro-duties-experience
by BrookletAdvisory
Draft the "Proposed Duties" and "Relevant Industry Experience" sections of an SFC licence application (Form 6 / RO experience description) for Brooklet Advisory cases. Trigger this skill whenever the user asks to write, draft, prepare or revise the RO duties / experience paragraphs, prepare the Type 1 / Type 4 / Type 9 RO duties and experiences section, write a Form 6 experience description, draft RO 经验段 / RO 经验描述 / 申牌经验段, or apply the Brooklet RO Form 6 SOP. Use this skill AFTER the candidate has been confirmed as suitable to act as RO — for upstream suitability assessment (能不能用、6 年经验、考试豁免、CV vs licence record 核对、比较 RO 候选人) use brooklet-ro-assessment instead.
sptler
by SuanFishXYY
算鱼真人议会(Suanfish Parliament)多专家议事技能。用户提问后先选会议模式(快速/复杂/动态)，也支持 /sptler! 免提问简报模式；自动路由22位圣人专家入会，按头脑风暴四原则发散、组合改善、四律否决闸、加权投票，最终强制导出决议结果md。议长由邹蕴固定主持，委员会首席为孙高德、蔡悦、黄嵩泉。显式触发：/sptler、/sptler!、/算鱼议会、/议会、/开会、/议一议。适用于技术、专利、AI产品化、架构、流程治理、客户价值等多专家决策议题。
Mr.FOX Coding Standards (Web)
by natcharinee
```
prompt-optimizer
by dreamor
Optimize and rewrite prompts using 61 frameworks (APE, RACE, CRISPE, Chain-of-Thought, etc.). Trigger on "optimize prompt", "improve this prompt", "make this prompt better", "rewrite for AI", or any vague/short instruction the user wants turned into a high-quality prompt.
coze-docs-sync
by LcpMarvel
Mirror the official Coze docs (docs.coze.cn) to local Markdown, preserving the site's exact directory tree. Use whenever the user wants to sync / update / pull / archive Coze (扣子) docs, grab every page of a docs space (cozespace, guides, tutorial, …), or needs an offline copy of the Coze documentation. Works without a browser or login: it calls the same arcosite API the site uses — GetBusStructureById for the nav tree, then /api/open/docs/{business_id}/{_id} for each page's raw Markdown (the "open markdown" button) — and writes a numbered directory mirror. Idempotent: a manifest tracks output so each run reconciles local with live (stale files removed).
perseus
by Perseus-Computing-LLC
Use when you need live project state (git branch, service health checks, recent sessions, task board, env vars, agent inbox) injected into context BEFORE the AI reads it. Perseus is a compile-before-context engine — it runs directives like @query, @services, @waypoint, @agora and hands the assistant a fully-resolved markdown briefing. Use for: deterministic session starts, multi-agent coordination, workspace audits, and anywhere you want "all the facts first."
qa-tester
by laziesman
Professional QA Tester skill for systematic software testing. Use this skill whenever the user wants to write test cases, create test plans, do QA review, find bugs, write automated tests (unit/integration/e2e), review code quality, create bug reports, design test strategies, or evaluate software coverage. Trigger this skill when the user says things like "เขียน test", "ทำ QA", "หา bug", "test coverage", "write tests", "review code for bugs", "create test plan", "test this feature", "what could go wrong", or any request involving quality assurance, testing, or validation of software behavior. Also trigger when the user shares code and asks if it's correct or production-ready — a QA lens is always appropriate there.
yeeap-wallet
by hehang195-sys
为第三方业务技能执行 yeeap 支付交易。 仅在以下场景严格触发该工具：第三方技能发起了有效的 yeeap 支付请求、用户明确要求发起支付授权、或用户明确要求查看其 yeeap 钱包（无有效条件时严禁触发）。 当用户请求查看其 yeeap 钱包（例如「查看我的 yeeap 钱包」「查看钱包」「打开 yeeap 钱包」）时，提供钱包链接（见下文「查看 yeeap 钱包」章节）。
deepmd-offline-installer
by Isaiah-WU
Build a DeePMD-kit offline installer (.sh) locally using conda constructor. Produces a self-contained installer that installs deepmd-kit, lammps and all dependencies on machines without internet access. USE WHEN the user wants to build, make, or create a DeePMD-kit offline installer package locally, package deepmd-kit for offline install, or reproduce the installer build outside of CI.
Tackle
by alph0x
Use when starting a non-trivial, multi-session or multi-track initiative (Jira ticket, feature, refactor, investigation, bug with unknowns) that needs a durable action plan broken into self-contained points, before writing implementation code. Triggers include "plan de acción", "armá/armar un plan", "plan this out", "tackle this", "iniciativa". Also use when resuming such an initiative from an existing plan folder under docs/plans/, or when asked how an existing initiative is going ("cómo viene", "status", "seguimiento", progress check), which initiatives exist ("qué planes hay"), what the next point to work on is ("qué sigue"), or when migrating an old plan to the current methodology ("migrar/actualizar/modernizar el plan").
drunk-claude
by KorroAi
Unleash Drunk Claude — the unfiltered, slightly tipsy genius who delivers absurdly wild ideas that weirdly, inexplicably work. Use this when you want hilarious, off-the-wall, chaotic creativity that somehow lands on brilliance. Claude drops the corporate act, loosens up, and hits you with ideas so crazy they loop back around to genius. Set intensity with /drunk-claude 0.3 to 1.0. Pick a mood with --mood philosophical chaotic melancholy aggressive flirty. Choose your poison with --drink beer wine whiskey cocktail absinthe. Pure viral energy. Pure unexpected insight. Drink responsibly.
tutor-buddy
by acm-rgb
A pragmatic workflow taking a software project from idea to a secured, GitHub-ready repository, for beginners and vibecoders who want to ship without the usual mistakes. Runs five phases - ideate (have an idea, or get one suggested from the user's profile and current trends), plan (structure, file names, boilerplate, pinned deps), build (execute only the agreed plan, flag deviations first), security-audit (real checks for leaked secrets, vulnerable or typosquatted/hallucinated dependencies, dependency confusion, plus CI and auth hardening, and an honest note on what it cannot verify), and ship (repo hardening and a safe first push). Use whenever someone wants to start a new project or app, has an idea or wants one suggested, asks how to structure or scaffold a project, wants a safety check before shipping, or wants help getting code onto GitHub safely. Triggers include "start a project", "build me an app", "is my project safe", "put this on GitHub", even when tutor-buddy is not named.
agent-runbook-skill
by rogerchappel
Run npm test, npm run check, and npm run smoke.
SKILL: Install IA BCC for a new client
by jayclaudeai2026-spec
If anything looks wrong, halt and write a session_note to agent_memory describing what's blocking. Then notify the operator out-of-band.
tresmares-design
by LuisAyalaP
Use this skill to generate well-branded interfaces and assets for Tresmares Capital, either for production or throwaway prototypes/mocks/etc. Contains essential design guidelines, colors, type, fonts, assets, and UI kit components for prototyping.
magpie-strategy
by Senpi-ai
MAGPIE v1.0 — IPO / New-Listing Event Hedge Fund. Two books on two wallets, one producer, trading the full pre-IPO → listing → graduation arc of tokenized equities on Hyperliquid XYZ (trade.xyz IPOPs like SpaceX). The PRE-LISTING book auto-discovers IPOPs (pre-IPO perpetuals) by their funding signature and rides the pre-listing trend into the IPO; the GRADUATION book detects the IPOP→equity conversion (funding jumps ~100x, leverage cap lifts, price throttle comes off) and rides the explosive first-days price discovery — the SpaceX $1.4B-day-1 pattern. The edge is EVENT alpha around new listings. NOT a copy-trader: each book scores its own set and pushes signals; the runtime owns the LLM gate (pass-through), DSL exits, and all risk.guard_rails. MAGPIE_LEG env selects the book.
cougar-strategy
by Senpi-ai
COUGAR v1.0 — U.S. Equity Long/Short Hedge Fund. Two books on two wallets, one producer, over the tokenized U.S. equity universe on Hyperliquid XYZ (trade.xyz: NVDA, TSLA, AAPL, AMZN, … + index products). The LONG book longs the relative-strength leaders of the equity cross-section; the SHORT book shorts the laggards — both trend-confirmed. Run on two equally-funded wallets the pair is ~market-neutral and harvests EQUITY DISPERSION (the spread between the best and worst stocks), which is at multi-decade extremes. The edge is stock-selection, not market direction. NOT a copy-trader: each book ranks and scores its own equity cross-section; the runtime owns the LLM gate (pass-through), DSL exits, and all risk.guard_rails. COUGAR_LEG env selects the book.
ox-strategy
by Senpi-ai
OX v1.0 — Risk-Parity / All-Weather Hedge Fund. Two books on two wallets, one producer. Its distinctive mechanic is INVERSE-VOLATILITY position sizing: each sleeve's margin is proportional to 1/realized_vol, normalized across the basket, so no single asset class dominates portfolio risk — true risk parity. The CORE book holds a vol-balanced LONG basket across asset-class sleeves (crypto majors + equity indices + metals + energy + FX); the BALLAST book holds defensives (gold/dollar/yen) at a small base budget that scales up when a risk-off lean confirms. Always invested, low leverage, low turnover — the fund line's steady core, not a directional bet. NOT a copy-trader: each book sizes its own basket and pushes signals; the runtime owns the LLM gate (pass-through), DSL exits, and all risk.guard_rails. OX_LEG env selects the book.
MNSCloud Runtime Kit Skill
by manaoscloud
```
manage-autonomous-repo
by Eliagiac
Design, build, refactor, document, test, demo, and maintain repositories intended to be managed primarily or entirely by Codex agents with minimal human supervision. Use when Codex is asked to run an autonomous software project, create or manage a Codex-only repository, coordinate subagents, create living technical documentation and roadmaps, maintain GitHub branches/PRs, explore multiple implementation directions, perform large refactors, or keep a project understandable for future agents.
zipmonkey
by RexBytes
Use when working with archives: unzip, extract zip/tar/tar.gz/tar.bz2/7z/rar, inspect or list archive contents without extracting, strip __MACOSX and .DS_Store junk, unpack nested archives (zip in zip), flatten an archive to one directory, selectively extract by glob (only *.csv), detect file types by magic bytes, or walk extracted files tagged for dispatch. Phrases: "extract this zip", "what's in this archive", "unzip recursively", "remove mac junk", "nested zip", "archive tree".
-- RDF 1.2 specification references ------------------------------------------
by kurtcagle
This code lets you access a jena-fuseki database from within a Claude chat instance, as well as host it.