AI & ML

arcanedge-ai Skill

Use when a task may need architecture, testing, access-control, design-system, API, release, data-model, subagent, or worktree-lifecycle reference documents. Helps select relevant docs, classify authority, and pass concise context to subagents.

arcanedge-ai Skill

Use for every repository task when Codex subagents are available. Makes the main agent the orchestrator, delegates bounded execution by default, enforces dependency-aware decomposition and fixed Luna/max routing, and requires main-agent verification.

ashhas Skill

Write or restructure an agent skill (SKILL.md). Use when creating a new skill, reviewing an existing one, or deciding whether something should be a skill at all. Covers naming, the description line, layout, formatting for agent readers, and the pre-publish checklist.

coinlobster Bundle

Build apps and agents on CoinLobster's crypto whale data, including the free keyless REST endpoints, the hosted MCP server, API keys and credit plans. Use when a developer asks how to integrate live whale trades, whale alerts, liquidations or the Smart Money Radar into their bot, app, dashboard or agent, or asks about the CoinLobster API or MCP server.

coinlobster Skill

The data picture around a crypto position. Use when the user is about to enter or exit, is holding a position and wants what to watch, asks if their side is crowded, what holding costs, whether whales are on their side, or wants their own signal cross-checked against smart-money flow. Reports state and recorded history with sample sizes; never advice, never a forecast. Keyless on every streamed pair: trade_context, whale_context and funding_matrix on MCP, or /api/public/baseline, /api/public/whale-context, whale trades and liquidation history over plain curl. Percentile judgment vs a pair's own recorded days is keyless too, and so are whale USD magnitudes over /api/public/whale-context. Depth (whale USD on the MCP tools, crowding, full 24h stats) needs an active developer plan (from $29/mo), and so does squeeze_score outright; liq_zones is keyless on the free pair only.

hklhaobin Bundle

Diagnose and fix Cursor Agent CLI on Termux/Android: empty Shell stdout with exit 0, no file side effects, unwritable /tmp, proot bind, cursor-termux patches, allowlist Shell(**), and MCP Bearer env empty causing 401 then SSE 405. Use when the user is on Termux, Android aarch64, agent/cursor-agent, cursor-agent-proot, or reports Shell fake-success.

hklhaobin Bundle

Bootstrap MEMORY.md, Agents.md, and an always-apply Cursor rule for Termux workspaces. Use when setting up a new Android/Termux device, when the user asks for a memory file, Agents.md, agent rules, or a reusable experience log, or when the workspace has no MEMORY.md yet.

hklhaobin Bundle

Bootstrap Cursor Agent CLI on Termux before it can run: official linux-arm64 node "cannot execute: required file not found", pkg mirrors all bad, dpkg lock/interrupted, openssl.cnf prompt, cursor-termux smoke test missing tree-sitter, npm allowScripts blocking native rebuild, agent vs cursor-agent musl MODULE_NOT_FOUND. Use when installing agent/cursor-termux on Android aarch64, or when `agent` fails before --help works.

huanyu-hibiki Skill

创始人 IP 战略决策 + 立项 skill，对应徐沪生《个人IP全流程拆解》第 1/2/9 章。 通过 7 问深度采访（一问一答、给候选）逼创始人想清楚三件事： (1) 为什么做个人 IP——时代背景 + 个人动机； (2) 创始人 IP ≠ 网红 IP——防 4 个错误观点、防邯郸学步； (3) 立项决心——不做"随便做做"、时间承诺、耐心周期、成功标准。 产出 strategy-memo.md（不可改的决策段 + 可追加的修订段）。 founder-ip 系统的入口 skill，建议第一个跑。其他子 skill（persona/business-model/funnel/opc-system）都引用此文档。 触发词："创始人IP战略"/"ip战略"/"ip-strategy"/"立项"/"起号"/"我该不该做IP"/" founder ip init"。

huanyu-hibiki Skill

OPC 内容生产流水线整合 skill，对应徐沪生《个人IP全流程拆解》第 8 章（OPC 改造版）。 徐沪生讲的是"编导+摄影+剪辑"的团队配置，对 OPC 不适用。本 skill 把它改造为"创始人 + AI + 自动化"的流水线。 整合现有工具链（cheat-on-content + 剪映 + Hyperframe/Remotion + Hermes Agent），定制 2 天 1 期的标准日程、AI 协作边界、瓶颈优化方案。 产出 opc-sop.md（持续优化）。 founder-ip 系统的收尾 skill，建议最后跑。触发词："OPC系统"/"ip-opc-system"/"生产流水线"/"2天1期怎么排"/"内容SOP"。

huanyu-hibiki Skill

创始人 IP 商业画布 + 咨询漏斗 skill，对应徐沪生《个人IP全流程拆解》第 4 章（创始人 IP 的商业价值）。 通过 5 问采访定制你的：商业模式画布、流量视频 vs 转化视频比例、咨询枢纽转化路径、私域沉淀流程、季度商业指标。 核心理念：徐沪生模式 = "IP 带咨询，咨询养产品，产品等 B 端机会"。忘掉卖货卖课，反而能卖得更好。 产出 business-model-canvas.md（季度更新）。 建议在 /ip-strategy + /ip-persona 之后跑。触发词："商业画布"/"ip-business-model"/"怎么变现"/"咨询漏斗"/"转化路径"。

shenseanchen Bundle

Rag Setup

shenseanchen Bundle

Rag Tools

shenseanchen Bundle

Rag Database

skillnerds Skill

Use when developing xskill itself — editing agent tool docstrings, changing the generate agent tool surface, or verifying what schema the Agno framework actually sends to the model. Covers the dump_schema workflow and where tool descriptions come from.

skillnerds Bundle

Use when installing, configuring, or operating xskill (the `xskill` CLI / `pip install xskill`) — starting the daemon, registering trajectory dirs, joining a team server, understanding how trajectories become Skills, or rebuilding/re-distilling the skill library after a model change.

takeoff69 Bundle

Generate structured, self-contained Codex task prompts. Use when the user asks for a Codex prompt, delegation task, fix, build, audit, ops/tuning, refactor, PR-series, or master-session brief to hand to another Codex run. Grounds every cited path and command, distinguishes directive from investigative work, and binds completion to explicit evidence.

takeoff69 Bundle

Run a self-distillation retrospective on recent OpenAI Codex sessions (or any Agent-Skills-compatible agent). Use when the user says 'retro,' 'retrospective,' 'distill,' 'what did we learn,' 'review sessions,' 'self-improve,' or after completing a major chunk of work. Covers three modes: quick retro (single session), deep distillation (batch review), and failure analysis. Outputs updates to skills, memories, AGENTS.md, and anti-pattern lists.

takeoff69 Bundle

Close a coding-agent session with durable repository state, proportional verification, documentation reconciliation, and a useful handoff. Use when the user says close out, wrap up, end session, session report, doc sync, preserve work, or asks for a final commit and push before stopping.

takeoff69 Bundle

Choose proportional verification and write evidence-bounded completion claims for coding-agent work. Use before saying work is fixed, done, working, passing, shipped, or deployed; when asked to verify a change; when relying on another agent's findings; or when reporting counts, artifact properties, UI behavior, live data, or production state.

takeoff69 Bundle

Operate Git safely when multiple humans or coding agents may share a repository, checkout, index, branches, or worktrees. Use before staging, committing, pushing, merging, rebasing, creating or removing worktrees, recovering apparently lost edits, handling staged files you did not touch, or preserving unfinished agent work under parallel-session contention.

tomosia-vietnam Skill

Run the open-pr e2e fixture through a real review in a fresh subagent, grade it against e2e/checklist.md with a second independent subagent, diagnose each failure to the prompt file that owns it, fix, and repeat until clean or the round budget runs out. Use when changing anything under src/ and you want evidence the review still behaves, not just green unit tests.

torvaldz Skill

Start and drive a large project (more than one agent session can hold) as a map of decision tickets on GitHub Issues, resolving one at a time until the way to the destination is clear. Self-contained (zero external-plugin dependencies: trailhead ships its own `trailhead-*` engine agents and needs only an authenticated gh CLI): map-based onboarding and a discuss→plan→execute→verify engine with atomic commits, with everything (map, discussion, plan, verification) living on the Issues. Invoke only when the user explicitly runs `/trailhead` (or a `/trailhead:<verb>` command) or asks to chart, adopt, or work a trailhead map, and do not auto-invoke for generic planning, project, or task-tracking requests.

star1342354 Bundle

评价全国大学生数学建模竞赛本科组的模型选择与算法方案。根据赛题、全部题目附件及按小问组织的方案，逐模型和算法联网查证，输出有依据的百分制诊断、具体问题、修改方案和优化脉络。逐问强制重读赛题原文并做四遍分析（抄读/自解/对照/反查），四遍记录是报告的必备部分。适用于建模方案审查、方法适配性评价和方案迭代；不要求完整论文或已完成实验。

veschin Bundle

Use when the user wants to design or research a subsystem, storage, or data flow inside an existing system ("спроектируй хранение", "design the storage", "сделай исследование", "run a research"), or to run development work through delegated workers, and demands no-fabrication discipline. Forces: ONE contract (ТЗ or a file created with the user) that all work strictly follows; every decision locked with the user BEFORE any action; every claim anchored to a verbatim quote - no quote, no statement, no action; scope limited to what the user asked. Hierarchy: the user is the CTO; this skill's agent is the architect under him with extremely low freedom; spawned workers are controlled autocomplete - template filling per work orders, refusing on anything underdetermined. Two domains: architect domain (TZ/QUOTES/REF/TODO/CODEX, guard-frozen) and development domain (research/impl/experiments/benches, worker-writable). Internal sub-skills - development (work orders, lanes, acceptance), calculator (sizing workbooks), dat

wgm66 Bundle

WPS 演示智能助手，通过自然语言操控 PPT，解决排版美化、内容生成、动画设置等痛点问题。适用于：幻灯片美化、内容生成、动画设置、母版编辑、批量处理。当用户提及 PPT、演示文稿、WPS 演示、WPP、幻灯片、美化时使用此 skill。

wgm66 Bundle

WPS 文字智能助手，通过自然语言操控 Word 文档，解决排版、格式、内容编辑、模板填写等痛点问题。适用于：文档排版、格式设置、目录生成、表格插入、样式管理、模板填写。当用户提及 Word、文档、WPS 文字、排版、目录、样式、填写、模板、表单时使用此 skill。

wgm66 Skill

WPS 文档校对专家，专注于文档的错别字检测、语病检查、格式一致性校对。当用户说'校对'、'审校'、'检查错别字'、'proofread'、'审阅'、'帮我检查文档'时使用此 skill。不处理排版、字体、表格插入、模板填写等 Word 编辑操作。

wizspec Bundle

Interrogate saved personas about a subject - cast each as an isolated agent with its own packet, collect findings and verdicts in a fixed vocabulary, confront disagreements, and synthesise the result. Use to judge whether a plan, document, diff, or decision holds up for the people it affects, or to ask those people questions in character; not for creating personas.

wowcube Skill

Use when verifying WowCube game code after a coder agent completes implementation. Contains two agent roles: Requirements Agent (checks completeness, GDD, regressions) and Template Agent (checks code against app_ai_template.h). The orchestrator deploys each agent separately.

yibie Bundle

将已经确认的多上下文设计记录成可交接、可受控修订的 SPEC。适用于工作需要跨会话、跨 agent 或需要保留设计决定时；不要用它替用户重新做决定。

yunhe-dev Skill

Validate and submit or update a CodexSkin theme from an Agent, anonymously for review or with an optional API key for immediate owned publishing.

yxy050208 Skill

Diagnose, configure, generate, simulate, and report on NI Multisim circuits through the Multisim MCP server. Use when the user asks to install or configure Multisim MCP, generate a Multisim schematic, run circuit simulation, analyze waveforms, or produce an electronic design report.

grp06 Bundle

Compile a completed Grill Me session, transcript, or decision discussion into a durable work item under `.agent/work/`: write `decision.md`, initialize `meta.json`, create an ExecPlan, run configurable `execplan-improve` passes, and invoke Goalcraft to activate a Codex `/goal` that executes the ExecPlan lifecycle. Use after grilling, when the user says "GrillCraft", "turn this grill into a goal", "create the ExecPlan and goal", "run Goalcraft after this grill", or asks for N ExecPlan improvements.

grp06 Bundle

Strategically audit an entire Codex thread, distinguish completed work from discussion, challenge its direction from first principles, and produce one paste-ready prompt for the main agent. Use when the user asks to reorient, reassess, audit, or get a task or thread back on track without continuing implementation or taking consequential action.

grp06 Skill

Read an existing ExecPlan, deeply analyze every referenced file and code path, and rewrite the plan with concrete, code-grounded improvements. Prefer work-item plans under `.agent/work/`, while preserving backward-compatible support for older singleton plan files.