AI & ML

hideki5123 Bundle

Operate Slack from the terminal via `slack-cli` (@urugus/slack-cli). Send messages, check unreads, search history, upload files, manage reactions/pins, schedule messages, and manage channels — all through Bash commands. Prefer this skill over `mcp__slack__*` MCP tools when you need: scheduling, unreads, search with pagination, reactions/pins, file uploads, multi-workspace profiles, canvas operations, or jq pipelines for filtered output. Trigger phrases include "slack-cli", "slack cli", "send slack message", "check slack unreads", "slack history", "search slack", "schedule slack message", "slack channel", "upload to slack", "slack reactions", "slack pins", "slack canvas", "slack users", "slack unread".

hideki5123 Bundle

Use confluence-cli to read, search, create, update, move, and delete Confluence pages and attachments from the terminal. Prefer this skill over `mcp__claude_ai_Atlassian__*` MCP tools when you need: page export with attachments, copy-tree, attachment management, profile-based multi-instance access, read-only mode, CQL search, or shell pipelines. Trigger phrases include "confluence", "confluence-cli", "confluence page", "read confluence", "search confluence", "create confluence page", "update confluence page", "confluence attachment", "confluence spaces", "export confluence", "confluence comments", "confluence copy-tree".

hideki5123 Bundle

Operate Notion from the terminal via the official Notion REST API on TypeScript + deno — search workspace, get/create/update/archive pages, query databases with filters and sorts, list/append/delete blocks, list and retrieve users, list databases, retrieve page comments. Two-tool stack: mise + deno (no node_modules, no pnpm, no bash). Token stays outside the agent: existence-only check, scoped `--allow-env=NOTION_TOKEN,NOTION_API_KEY,NOTION_TOKEN_FILE` and `--allow-net=api.notion.com` flags only, no `.env` files, never echo or print the token. Token resolved in order: NOTION_TOKEN → NOTION_API_KEY → contents of the file at NOTION_TOKEN_FILE (suits agenix / sops-nix / 1Password CLI mounts). Walks first-time users through creating a Notion Internal Integration and sharing pages with it. Trigger patterns (match any variation): notion / Notion / notion-cli / notion cli / notion api / notion page / notion pages / notion database / notion db / notion search / search notion / find in notion / create notion page / ne

danmackenz Skill

Diagnoses hidden token/cost leakage across a user's AI CLI tools (Claude Code, Codex CLI, Gemini CLI, and others) by checking hooks, MCP server health, plugin build integrity, project-level config drift, and cross-provider spend via ccusage — cross-referenced against the user's known-tools baseline from ghostspend-setup. Use this whenever a user reports unexpectedly high AI tooling spend, wants a routine health check, or asks about background processes / hooks / MCP servers / other CLIs silently consuming tokens.

univa-agent Skill

Codex UniVA Media Ops

univa-agent Skill

Claude Code UniVA Media Ops

univa-agent Skill

UniVA External Agent Bridge

dev-ansh-r Bundle

Convert an ONNX model to a QNN model library for Qualcomm HTP using the classic flow - static-shape preparation, calibration input_list generation, qnn-onnx-converter with W8A16 quantization, then qnn-model-lib-generator to produce an aarch64 .so, validated with qnn-net-run. Accepts AIMET .encodings via --quantization_overrides alongside --input_list. Use for Qualcomm/Hexagon HTP model conversion, whether a model will run on the NPU, FP16 or float_fallback errors, or any qnn-onnx-converter, model-lib-generator, input_list, bias_bw or W8A16 question. The DLC alternative is qnn-context-binary.

dev-ansh-r Skill

Quantize an ONNX model for Qualcomm HTP with AIMET - QuantizationSimModel (quantsim) calibration, AdaRound adaptive weight rounding, cross-layer equalization, bias correction and automatic mixed precision - exporting a .encodings file consumed by qnn-onnx-converter or qairt-converter via --quantization_overrides. Use for post-training quantization (PTQ), W8A16/W8A8 accuracy recovery, per-layer quantization sensitivity analysis, or when a converted model runs but its accuracy dropped. Runs on the AIMET host, not the board.

mike22890 Bundle

多 agent 对抗审查与高置信度验证。当需要多 agent 协作审查、交叉验证、对抗性测试、代码安全审计、漏洞狩猎、红队/蓝队演练、高置信度结论获取、复杂系统验证、多方评审、自动化 QA 流程时使用。

hideki5123 Bundle

Call the OpenAI Platform API from the terminal via TypeScript on deno — chat completions, Responses API, reasoning (o-series), embeddings, image generation, audio (Whisper transcription, TTS), files, batches, moderation, and model listing. Two-tool stack: mise + deno (no node_modules, no pnpm, no bash). Models are resolved dynamically per session against `client.models.list()` — never hardcoded — and any detected upgrade prompts the user via AskUserQuestion (even in auto-mode). Secrets stay outside the agent: existence-only key checks, scoped `--allow-env=OPENAI_API_KEY` and `--allow-net=api.openai.com` flags, no `.env` files. Scoped to the Platform API specifically — for open-ended chat/conversation ("ask gpt", "chat with gpt") prefer the codex-server skill instead, which uses the user's ChatGPT subscription rather than API-key billing. Use this skill only when the request is explicitly about the Platform API surface (embeddings, image generation, audio, moderation, batches, files, or a chat/Responses call m

hideki5123 Bundle

Agent Swarm

hideki5123 Bundle

Generate AGENTS.md (cross-tool agent context — Codex CLI, Aider, Cursor, etc.) and wire CLAUDE.md as a symlink so Claude Code reads the same file. Delegates content generation to Claude's built-in /init when needed, then renames + symlinks. Idempotent. Triggers — /agents-init, "generate agents.md", "create agents.md", "share claude and codex context", "set up agents file", "init agents", "agents-init", "cross-tool agent context", "wire claude and agents".

hideki5123 Bundle

Generate a persistent PROJECT-KNOWLEDGE.md profile for any codebase that gives subagents instant deep domain expertise. Analyzes project structure, architecture, code conventions, tech stack, data flow, and domain concepts. The profile is designed to be loaded into Agent tool prompts so subagents can answer deep codebase questions like an expert teammate. Use when the user asks to profile a project, generate project knowledge, create a codebase profile, build a context pack, analyze a codebase for subagents, or says "profile this project". Trigger phrases include "profile project", "generate project knowledge", "codebase profile", "subagent context", "knowledge profile", "project profile", "profile this codebase", "build project knowledge", "create context pack", "generate codebase knowledge", "index this project", "make subagent", "generate subagent", "deep knowledge", "project expert".

hideki5123 Skill

Sync a session-recap directory (summary.md + note.md + meta.json) to Notion under the explicit "My Agent Notes / Session Recaps" database via the notion-cli skill. The Notion target is fixed by design — the value of this skill is being explicit about its destination, not configurable. Appends to the shared session-recap manifest on success and retries previously-failed uploads on next invocation. Cleanup and retention are owned by session-recap, not this skill. Standalone — pass a recap dir path, or invoke in retry-only mode. Auto-invoked by session-recap via the chain in its ~/.claude/session-recaps/.recap-config.json. Use when the user says "recap to notion" / "Notionに recap を上げて" / "recap upload" / "/recap-to-notion" / "recap retry" / "再アップロード" / "Notion同期".

hideki5123 Skill

Install a PreCompact hook that auto-generates .agent/local/HANDOVER.md before context compaction, plus a SessionStart hook that restores context from a previous handover. Run once to install, then it works automatically forever. Use when the user says "handover", "setup handover", "install handover hook", or wants seamless context transfer between sessions.

hideki5123 Bundle

Run a multi-model adversarial mutual review with facilitator synthesis: 2+ panelists (different model families where possible — use codex-server / codex-cli for the external-family seat) independently answer, adversarially critique each other across rounds, and the main session acts as facilitator, synthesizing agreements, live disagreements, and calibrated confidence into one answer. Use ONLY on high-stakes, hard-to-reverse claims, designs or decisions, and only when the user EXPLICITLY asks to debate or adversarially review with multiple models — "discuss with Opus and Codex", "red-team this", "devil's advocate", 敵対的レビュー, 相互レビュー, 多モデル討論, 議論させて. A full run costs ~6 subagent/CLI dispatches: state the panel size and round count and get a go-ahead before launching. Where the Workflow tool is available, run the panel as a Workflow (judge-panel / adversarial-verify) instead of ad-hoc fan-out. Do NOT self-trigger on a bare "are you sure?" — that is a cue to OFFER the panel, not to run it.

hideki5123 Bundle

Fans out multiple agents in parallel via the Workflow tool to investigate a codebase question — bug root cause tracing, cross-file or cross-repository behavior tracing, "why does X happen for some cases but not others" questions — then synthesizes a ranked, file:line-cited answer in chat. The point of this skill is explicitness: it announces the investigation dimensions and agent count before launching, narrates progress, and reports a structured synthesis, rather than leaving multi-agent orchestration as an implicit background behavior the user can't see into. Supports investigations that span more than one local repository on disk (e.g. a backend service and a client app that talks to it). Investigation only — does not edit code or fix anything; pair with another skill or a follow-up request for that. Use when the user wants to investigate a bug, trace behavior across files or repositories, understand why something happens under some conditions but not others, or explicitly asks for a parallel/multi-agent i

tomo214 Skill

Lean-context working rules for Claude Code — size the task, reuse what is known, search before reading, read only what is needed, cap tool output, debug by hypothesis, escalate instead of looping, stop once verified. Never trades away tests, security checks or completeness. Use for any non-trivial coding, build, debug or research task, in large repos or with big files/logs, before spawning subagents, when fixes keep failing, and when the user mentions tokens, context, limits, cost, effort, model choice, Kontext or "sparsam".

sergekostenchuk Bundle

Use when a public website needs an AI assistant with source-bounded answers, RAG/content scope, injection protection, escalation, logging, rate limits, refusals, and no hallucinated product/legal/price claims. Planning or implementation handoff only; do not expose secrets or give destructive tools.

dcidda Bundle

Réglementation française applicable aux intermédiaires en assurance (IAS) et en opérations de banque et services de paiement (IOBSP) immatriculés à l'ORIAS : catégories et formalités ORIAS, capacité et honorabilité professionnelles, associations agréées, RC professionnelle et garantie financière, formation continue (DDA 15 h, crédit immobilier 7 h, IOBSP sans activité de crédit immobilier), formation à distance, rémunération, réclamations, LCB-FT et responsabilité du mandant. Utiliser dès qu'une question porte sur les obligations réglementaires d'un courtier, agent général, mandataire d'assurance ou intermédiaire en crédit en France. Ne couvre pas les autres professions dites de courtage : agents immobiliers relevant de la loi Hoguet, courtiers en travaux, en énergie ou en marchandises.

sergekostenchuk Bundle

Tactical LLM-friendly site optimization for making a website the best citable answer for AI assistants and RAG retrieval. Use for auditing and optimizing sites such as mlllm.io for llms.txt, AI citation readiness, pillar pages, direct-answer blocks, TL;DR sections, FAQ/schema, article/news schema, topic-to-URL matrices, external citation signals, weekly LLM citation monitoring, and content architecture for ChatGPT, Claude, Perplexity, Gemini, Copilot, Google AI features, Bing, Brave, and other assistant/search retrieval systems.

sergekostenchuk Bundle

Use when the task involves a filesystem-based LLM Wiki or Obsidian vault, especially Ar9av/obsidian-wiki projects using .skills, .env, ~/.obsidian-wiki/config, AGENTS.md, .manifest.json, index.md/log.md/hot.md, ingest/query/lint workflows, or Russian prompts like "добавь в вики", "сделай ingest", "lint wiki", "обнови index/log", or "разложи по obsidian-wiki".

sergekostenchuk Bundle

Plan and record evidence-backed LLM and AI-search citation checks across approved assistant/search surfaces with timestamped queries, locales, modes, cited URLs, snippets, screenshots or transcripts, and caveats. Use this skill when the user asks whether ChatGPT, Perplexity, Claude, Gemini, Copilot, or another assistant cites a target site, but do not claim citation without direct evidence.

aa0101181514 Bundle

Retrieve real Taiwan court judgments with verifiable citations before answering any question about Taiwan law or case law. Use when the user asks about 台灣/臺灣 judgments, 判決, 法院見解, legal precedents, or wants case support for a legal argument. Wraps the twlegalrag CLI (pack subcommand); answers must cite only citation_ids present in the returned bundle.

farzammohammadi Bundle

Executes an approved plan by ALWAYS orchestrating the build through focused agent workflows, then taking the role of orchestrator and co-owner to independently verify every change, hunt gaps, and fix whatever was missed until the work is fully home — no gaps, bugs, or loose ends. This is the "I" (Implement) of RRPIR: it runs after create-plan and before the optional review. Use this skill whenever it's time to build, implement, or execute an approved plan — when the user says "implement this", "build it", "execute the plan", "do the work", "make it happen", or "let's build it", and especially right after a plan has been approved. Reach for it for any multi-step implementation rather than hand-coding everything inline. Note: verification here means confirming the orchestrated build delivered the plan without gaps — it does NOT run the separate /review quality gate (coverage analysis, bug-hunt, MR prep), which the user runs on their own afterward.

framecoreworks Bundle

Use this skill for provider-neutral character design systems, identity anchors, expression sheets, outfit variants, consistency rules, and prompt handoffs.

farzammohammadi Skill

Take on a substantial, multi-step effort as a strategic orchestrator — intake the goal, research and plan it solo, execute it through parallel/sequential agent workflows, then personally verify the result end to end. Use this whenever the user wants to tackle something big or ambitious: a large feature, a cross-cutting refactor, a migration, a codebase-wide audit, a "do all of X / sync everything / overhaul Y" task, or any effort that won't fit in one pass and needs to be planned and orchestrated with rigor. Trigger when the user invokes it by name, says things like "orchestrate this / run a workflow / use subagents / plan and build this / take this on end to end," OR simply describes a meaty multi-part task and wants it done thoroughly and efficiently — even if they never say the words "workflow" or "orchestrate." Do NOT use it for small, single-step tasks that one pass handles directly.

sebas0430 Skill

Get an adversarial, independent evaluation of a decision, document section, or trade-off before committing to it. Dispatches to the decision-critic agent in a fresh, isolated context so it evaluates without inheriting this conversation's reasoning or attachment to prior choices. Use when the user types /decision-critic, or asks for a hard/critical/skeptical review of something already decided or drafted.

s2terminal Skill

markdown（.mdファイル）を新規作成・編集・添削するときに必ず使う。README、ドキュメント、仕様書、SKILL.md、CLAUDE.mdなどあらゆるmarkdownの執筆・修正で起動し、LLMが解釈しやすいmarkdownを出力する。「markdownを書いて」「ドキュメントを作って／直して」「READMEを更新」「.mdを編集」などmarkdownの記述が関わる作業すべてが対象。空間的レイアウトに依存せず線形構造と明示的な言葉で表現する。

framecoreworks Bundle

Use this skill to create final provider-neutral image prompt packs from approved brief, references, direction, and copy.

framecoreworks Bundle

Use this skill to create final provider-neutral video prompt packs from approved storyboard, motion direction, references, and copy.

hongbozcode Bundle

设计或扩展可复用的 Python AI 后端模板，覆盖 Core、业务集成、RAG、Agent 和工作流；适用于架构落地与模块解耦。

yimao1234 Bundle

3D 次世代国漫打斗提示词引擎。把任意题材的打斗（古武/仙侠/现代/都市/科幻/赛博/机甲/校园）蒸馏重写为可直接投喂视频大模型的「国漫级动作提示词」。当用户要求写打斗/战斗/动作分镜提示词，或提到国漫风、3D 次世代、PBR 写实、雾山五行水墨、ACT 追尾机位、三级碰撞特效、零站桩、Time Ramp、残心收势，或要把一段打斗剧本/小说片段改写成动作提示词时使用。

brett-fisher-research Skill

Use this skill to source contacts for a specified ideal customer profile using the Apollo MCP server. Only search using Apollo endpoints that do not consume any credits.

xxjrq Bundle

自动完成 Android 应用的 Google Play 上架工作。Use when an agent needs to inspect an Android project, build or locate a release AAB, upload binaries and store assets, fill Play Console listing and policy forms, create a Production release, send changes for review, check review or publication status, or read a rejection and fix the app or listing. Requires Easy WebBridge and an already logged-in Google Play Console browser.