AI & ML

heyzude Bundle

Use when coordinating 2+ specialized Codex sub-agents on one objective, especially for parallel workstreams, competing implementation options, or split implementation and review lanes that need explicit handoffs.

ruan-cat Skill

当用户要求在 bug 已经定位并修复后，记录排错经验、事故结论、AI 记忆更新、 复盘摘要或本地 MCP 记忆时使用。这个技能只负责沉淀 "发生了什么、为什么会发生、如何修好、以后要记住什么"， 不要把它用于实际修复 bug。

aradotso Skill

Context layer for AI data agents - query warehouses accurately with semantic layers, metrics, and wiki knowledge through MCP

aradotso Skill

Context layer for AI data agents - teach Claude Code, Codex, and AI agents to query data warehouses accurately with semantic layer, wiki knowledge, and MCP tools

aradotso Skill

An executable context layer that teaches AI agents to query data warehouses accurately through MCP with skills, memory, and a semantic layer

emilfihlman Bundle

Manage explicit GNU Screen targets that recover registered Codex work after usage exhaustion clears or the exact model-capacity warning appears. Supports goal resume, literal Continue, or both. Use only when the user explicitly requests this screen-based automation or asks to inspect, configure, register, unregister, start, stop, or remove its state. Do not use for generic process keepalive or unrelated long-running work.

emilfihlman Bundle

Manage explicit GNU Screen targets that recover registered Codex work after usage exhaustion clears or the exact model-capacity warning appears. Supports goal resume, literal Continue, or both. Use only when the user explicitly requests this screen-based automation or asks to inspect, configure, register, unregister, start, stop, or remove its state. Do not use for generic process keepalive or unrelated long-running work.

firecrawl Skill

AI-powered autonomous data extraction that navigates complex sites and returns structured JSON. Use this skill when the user wants structured data from websites, needs to extract pricing tiers, product listings, directory entries, or any data as JSON with a schema. Triggers on "extract structured data", "get all the products", "pull pricing info", "extract as JSON", or when the user provides a JSON schema for website data. More powerful than simple scraping for multi-page structured extraction.

firecrawl Skill

Extract clean markdown from any URL, including JavaScript-rendered SPAs. Use this skill whenever the user provides a URL and wants its content, says "scrape", "grab", "fetch", "pull", "get the page", "extract from this URL", or "read this webpage". Handles JS-rendered pages, multiple concurrent URLs, and returns LLM-optimized markdown. Use this instead of WebFetch for any webpage content extraction.

firecrawl Skill

Web search with full page content extraction. Use this skill whenever the user asks to search the web, find articles, research a topic, look something up, find recent news, discover sources, or says "search for", "find me", "look up", "what are people saying about", or "find articles about". Returns real search results with optional full-page markdown — not just snippets. Provides capabilities beyond default agent web search tools.

fox1245 Skill

Dense vector embedding storage and cosine search via sqlite-vector adapter.

giordanorogers Bundle

Run active learning sessions that diagnose understanding, teach from first principles, build a reusable mental model, quiz, give feedback, correct misconceptions, create transfer practice, and produce durable follow-up. Use when the user says teach me, help me understand, learning session, tutor me, quiz me, Socratic, practice, explain then test me, learn this repo, learn this paper, or asks to build real understanding rather than just receive an answer.

giordanorogers Bundle

Build and maintain sourced Obsidian/Markdown knowledge bases with raw-source ingest, source notes, compiled concept/wiki notes, links, frontmatter, answer-from-vault workflows, and vault health checks. Use when the user mentions Obsidian, vault, wiki, LLM Wiki, knowledge base, second brain, source ingest, compile these sources, save this to notes, answer from my vault, backlinks, frontmatter, note health, or Markdown knowledge management.

hassancs91 Bundle

Generates a consistent illustrated image for every scene of a story — keeping characters, locations, and visual style coherent across the whole sequence using cascading reference images. Reads a {slug}_scenes.json (from scene-splitter), proposes a visual style, ASKS the user for the aspect ratio and image model, extracts a visual bible, drafts a continuity-classified shot list, generates one reference image per character + location, then one image per scene (with a strict no-text/no-border constraint), and records every image in {slug}_images.json. Images are generated through the Fal API via the bundled fal_image.py script, with a pluggable model registry (nano-banana-2, nano-banana-pro, seedream-4, or any Fal model you add). Use this skill whenever the user wants to illustrate a story, generate scene images, draw the pictures for a storybook, build the illustrations, or produce a consistent visual sequence. Trigger on phrases like "illustrate this story", "generate the images", "draw the scenes", "make the

hassancs91 Bundle

Generate PNG images by building a real Three.js 3D scene and capturing one frame headlessly — no image model involved. Takes a text prompt and an aspect ratio (1:1, 16:9, or 9:16) and produces a rendered PNG. Use this skill whenever the user asks to generate, create, or render an image, picture, scene, wallpaper, thumbnail background, or product shot with this system — especially with phrases like "generate an image of", "render a scene", "create a PNG", "no image model", "3D render of", or a prompt paired with an aspect ratio. Also use it when the user wants an image in a named visual style (dark studio, Apple light, nature, sunset, night, underwater). Not for editing existing images, 2D vector art, or animated video (VidTSX handles video).

hongikarchi Bundle

Prepare the Rhino model for D5 Render — optionally strip block/point/curve/hatch/dim, and give every un-mapped object a texture mapping (copy from an already-mapped object, or a box 1000×1000×1000 mapping on a shared origin) so materials render consistently. D5's actual push is a manual click in the D5 panel (no scriptable command). Asks whether to work on a copy, dry-runs first, then applies. Trigger when the user wants to export/prep/hand off to D5 / D5 Render / D5 Sync / LiveSync, or "match mapping" for render. Fire on Korean or English requests, e.g. "d5로 내보내기", "d5 렌더 준비", "매핑 맞춰줘", "박스 매핑", "match mapping", "export to d5", "prep for d5 render".

hongikarchi Bundle

Remove duplicate / overlapping Rhino curves that SelDup misses — curves tracing the same path but stored as different object types (LineCurve vs NurbsCurve vs ArcCurve), plus shorter sub-segments lying on longer ones. Runs a non-destructive diagnose+dry-run via Rhino MCP, reports exact-dup and partial-overlap counts, confirms scope with the user, then deletes. Trigger when the user wants to clean / dedup / remove duplicate or similar curves or lines on the current Rhino selection. Also fire on natural-language requests in Korean or English, e.g. "중복 곡선 정리해줘", "겹친 선/커브 지워줘", "비슷한 선 정리", "중복선 삭제", "clean similar curves", "remove duplicate lines/curves", "dedup curves".

ix-infrastructure Skill

Flag parsing

ix-infrastructure Skill

Argument parsing

jonnybottles Bundle

Create, edit, publish, and manage Microsoft Copilot Studio agents entirely from the command line using the Power Platform CLI (`pac copilot`) plus the Dataverse Web API, with no Copilot Studio web UI and no browser automation. Covers scaffolding a live agent in one step, cloning an existing agent to a local YAML workspace, the pull/push sync loop, publishing, controlling authentication mode and access control, verifying that an agent is private to its owner, and testing an agent conversationally via Direct Line or the authenticated Copilot Studio conversation API. Use whenever the user asks to create a Copilot Studio agent, build an agent from the CLI, clone/edit/push/publish an agent, change agent auth or sharing, script agent creation, or test a Copilot Studio agent. Triggers: create copilot studio agent, pac copilot, copilot studio cli, agent as code, clone agent, push agent, publish agent, copilot studio yaml, agent definition, copilot studio automation, share agent, private agent.

jonnybottles Bundle

Attach a remote MCP (Model Context Protocol) server to a Microsoft Copilot Studio agent entirely from the command line - no Copilot Studio web UI, no browser automation. Covers authoring the OpenAPI/swagger definition with x-ms-agentic-protocol mcp-streamable-1.0, creating the Power Platform custom connector with `pac connector create`, creating a no-auth connection via the Power Apps REST API, creating the required Dataverse connectionreference row, and writing the TaskDialog / InvokeExternalAgentTaskAction / ModelContextProtocolMetadata action YAML into the agent workspace so `pac copilot push` succeeds. Also covers restricting which MCP tools the agent may call. Use when the user wants to add an MCP server or MCP tool to a Copilot Studio agent, wire up a custom MCP connector, or script MCP tool onboarding. Triggers: add mcp server to copilot studio, mcp tool copilot studio, mcp custom connector, x-ms-agentic-protocol, copilot studio mcp, streamable mcp connector.

jonnybottles Skill

Create, attach, troubleshoot, and verify Copilot Studio agent flows from the CLI and APIs. Use this skill whenever the user asks for an agent flow, "When an agent calls the flow", Power Automate tool, InvokeFlowTaskAction, AI Builder action, flow input/output, or a flow that writes to Dataverse - even if they do not name the skill. Covers solution-aware workflow clientdata, embedded connector bindings, Maker-mode tools, publishing, and proof that the flow actually ran.

jonnybottles Bundle

Verify a Microsoft Copilot Studio agent actually works at runtime by holding a real conversation with it from the terminal, and prove that a specific tool (MCP server, connector, or knowledge source) genuinely fired rather than the model answering from its own knowledge. Generates a ready-to-run Node test script that performs an Entra ID device-code sign-in and talks to the authenticated Copilot Studio conversation API via Microsoft's official agents-copilotstudio- client, including the one-time app registration that grants CopilotStudio.Copilots.Invoke. Also diagnoses common failure modes: connection-manager blocking, Integrated-auth channel errors, and empty responses. Use when the user wants to test, verify, smoke-test or chat with a Copilot Studio agent from the CLI, confirm a tool is being called, or debug an agent that returns nothing. Triggers: test my copilot studio agent, did the tool fire, agent returns nothing, smoke test agent.

jonnybottles Skill

Test Copilot Studio agents from the terminal or authenticated portal and prove a tool really fired. Use this skill whenever the user asks to test, smoke-test, invoke, or debug an agent; verify a flow/connector/MCP tool call; diagnose plausible but false agent answers; handle CopilotStudio.Copilots.Invoke device-code auth; or inspect flow runs and Dataverse side effects - even if they do not name the skill. Includes a Playwright fallback for unattended CLI authentication.

koreaben777 Bundle

Route Aside Browser work between its Codex MCP tool and direct CLI for interactive tabs, repeatable browser flows, or MCP transport failures.

koreaben777 Bundle

Use when the user asks what the other coding agent (Claude Code) has done in this project since last time, or asks to detect, catch up on, sync, or turn the pair agent's work into shared knowledge - reading both the project files and the pair agent's own chat threads, then writing a dated pair-sync note that separates verified facts from claims. Triggers include "클로드가 뭐 했는지", "페어 에이전트 작업 탐지", "지난번 이후 변경 지식화", "pair sync", "catch up on claude".

koreaben777 Bundle

Safely check and update GitHub-installed third-party Codex plugins and local skills. Use when the user asks to inspect, update, preserve, or automate updates for FableCodex, Ponytail, Archify, Graphify, Open Code Review, SkillOpt-Sleep, agency-router, codebase-memory-mcp, Superpowers upstream, or other third-party Codex skills/plugins without overwriting local custom work.

lobu-ai Skill

Use Lobu MCP for shared, permission-aware company context, durable organizational memory, and governed actions. Trigger when the user asks what the organization knows, needs context from connected company systems, wants to preserve a durable fact or decision, or needs to discover and use Lobu SDK capabilities.

lobu-ai Bundle

Use when working inside a Lobu project generated by @lobu/cli or any repository centered on lobu.config.ts, AGENTS.md, agent prompt files, local skills, and evals. This skill helps a coding agent inspect the right files, make Lobu-native changes, keep the stack runnable, and validate semantics with chat tests and evals.

whisperas Bundle

Default browser automation workflow for this machine using Playwright CLI and the local wrapper scripts in this skill. Use when a task involves opening websites, navigating pages, clicking UI elements, filling forms, taking screenshots, collecting page data, handing login back to the user, or downloading files, and prefer Playwright over browser MCP flows or ad hoc browser commands.

philsaxton Bundle

Set up an agent workspace (the garage) around independent application repositories, including local scratch and worktree locations, or reorganize existing work to separate application code, generated artifacts, and agent support. Use for workspace setup and domain separation, not routine feature work or Git branch/worktree cleanup.

squerne Skill

Plan and draft LinkedIn posts optimized for 360Brew, LinkedIn's decoder-based ranking model. Use when the user wants LinkedIn post ideas, a content plan, a post draft, or asks why their LinkedIn posts get no reach. Grounds every post in the user's real stories and enforces the 2026 ranking rules (saves over likes, no hashtags, analytical first line, pillar discipline).

ton-org Skill

Use this skill when invoking TON MCP tools directly from the command line. It calls any TON wallet tool via `npx @ton/mcp@alpha <tool_name> [--arg value ...]` without starting an MCP server session. Also use when querying wallet info, checking balances, sending transactions, or running raw TON MCP commands from the terminal.

ton-org Skill

Buy or sell Backed xStocks (tokenized equities, e.g. TSLAx) on TON using jetton addresses from the xStocks API and TON MCP swap tools. Use when the user asks for xStocks, xStock, TSLAx, stock tokens on TON, or buying/selling equity jettons via Omniston.

ton-org Skill

Use this skill when creating or deploying a TON agentic wallet. It generates operator keys and deploys an on-chain agentic wallet. Also use when setting up a new agent wallet, onboarding a wallet, or when any wallet operation fails because no wallet is configured. This skill is a prerequisite before sending, swapping, or managing assets on TON.

stackhawk Bundle

Use this skill when a user or agent needs to query the StackHawk platform for security reporting, findings analysis, or app management. Triggers include: "stackhawk api", "security posture", "findings report", "show me findings", "untriaged findings", "which apps", "scan history", "security dashboard", "triage", "what needs attention". Uses the hawk CLI (`hawk op` subcommand) for all platform queries. Do NOT use for running scans (use the hawkscan skill for "scan my app", "hawkscan", "stackhawk.yml", "DAST") or for fixing/remediating code or vulnerabilities — this skill only reads and reports platform data.