后端与 API
后端服务、API、Webhook 和服务端工具
Skills 列表
blave-quant
Use for: (1) Blave market alpha data — 籌碼集中度 Holder Concentration, 多空力道 Taker Intensity, 巨鯨警報 Whale Hunter, 擠壓動能 Squeeze Momentum, 市場方向 Market Direction, 資金稀缺 Capital Shortage, 板塊輪動 Sector Rotation, Blave頂尖交易員 Top Trader Exposure, kline, alpha table, 市場情緒 Market Sentiment, screener saved conditions, Hyperliquid top trader tracking (leaderboard, positions, history, performance, bucket stats), Taiwan stock daily OHLCV, forward-adjusted prices, institutional investor buy/sell, margin trading data, shareholding distribution, quarterly fundamental statements — income statement, balance sheet, cash flow, and broker/dealer daily buy/sell by branch (台股日K/向後調整/三大法人/融資融券/股權持股分級表/綜合損益表/資產負債表/現金流量表/分點買賣超); (2) CME / ICE futures OHLCV — WTI crude oil (CL), gold (GC), Brent crude (BRN); daily/hourly/minute candles from 2010; (3) Taiwan Futures OHLCV — TXF (台指期近月連續); daily/intraday candles (1d/1m/5m/15m/30m/60m), 1d from 2013-12-30 and intraday from 2014-01-02; (4) BitMart futures/contract trading — opening/closing positions, leverage, plan orders, TP/SL, trailing stops, account management, sub-account transfers; (5) BitMart spot trading — buy/sell, limit/market orders, account balance, order history, sub-account transfers; (6) OKX trading — spot and perpetual swap, order placement, positions, balance; (7) Bybit trading — spot and derivatives/perpetual swap, order placement, positions, balance, TP/SL; (8) BingX trading — spot and perpetual swap, order placement, position management, leverage, TWAP orders, OCO orders; (9) Bitget trading — spot and futures, order placement, position management, leverage, plan orders; (10) Binance trading — spot and USDS-M futures, order placement, positions, leverage, algo orders, OCO/OTO/OTOCO; (11) Bitfinex trading & funding — spot, margin, funding/lending (submit offers, loans, credits), wallet transfers; (12) KuCoin trading — spot and futures/perpetual contracts, order placement, position management, leverage, stop orders, account management; (13) Taiwan stock lookup/quote/PE — look up Taiwan stock codes and company names, daily quotes (open/high/low/close, volume), PE ratio, dividend yield, PB ratio for listed (上市) and OTC (上櫃) stocks: use the Blave API endpoints (`studio/market/twstock/list`, `/info/<stock_id>`, `/price/<stock_id>`, `/quote/<stock_id>`, `/per/<stock_id>`) — NOT the raw TWSE/TPEX open API. The raw TWSE/TPEX open API (no key required) is ONLY for two things Blave has no endpoint for: trading-halt status and a one-shot full-market PE/yield/PB scan; (14) 台股分點買賣超 — search broker branch code by name (`broker/search?name=`), then query by stock (`broker/stock/<stock_id>?date=`) or by broker branch (`broker/trader/<trader_id>?date=`); single-day per request, loop for multi-day; via Blave API; no CAPTCHA required; (15) Gate.io trading — spot and USDT-settled perpetual futures, order placement, positions, leverage.
blave-tw
arize-instrumentation
Adds Arize AX tracing to an LLM application for the first time. Detects the stack, routes to the single matching integration doc, wires auto-instrumentation after user confirmation, and verifies traces land. Use when the user wants to instrument their app, add tracing from scratch, set up LLM observability, integrate OpenTelemetry or openinference, or get started with Arize tracing.
arize-ai
lark-mcp
飞书/Lark 官方 MCP 集成。支持发送消息、创建群组、操作多维表格（Bitable）、导入/搜索文档、知识库查询。触发词：飞书、Feishu、Lark、多维表格、bitable、飞书文档、飞书群。
whatevertogo
trigger-tasks
Build AI agents, workflows and durable background tasks with Trigger.dev. Use when creating tasks, triggering jobs, handling retries, scheduling cron jobs, or implementing queues and concurrency control.
triggerdotdevtrigger-setup
Set up Trigger.dev in your project. Use when adding Trigger.dev for the first time, creating trigger.config.ts, or initializing the trigger directory.
triggerdotdev
just-scrape
通过 ScrapeGraph AI CLI 搜索、抓取、爬取、提取结构化数据以及监控网页。当用户要求搜索网络、抓取网页、从 URL 获取内容、从网站提取 JSON、爬取文档或网站部分、监控页面变化、检查请求历史、查看 ScrapeGraph 积分或验证 API 设置时使用。
scrapegraphai
inngest-events
Use when designing event-driven workflows, decoupling services, implementing fan-out patterns (one trigger, many downstream handlers), implementing idempotent event handling with IDs (24-hour dedupe window), or handling at-least-once delivery from external sources like Stripe webhooks. Covers Inngest event schema, payload format, naming conventions, IDs for idempotency, the ts param, fan-out patterns, and system events like inngest/function.failed.
inngestinngest-durable-functions
Use when building functions that must survive process crashes, retry automatically on failure, run on a schedule, react to events, or maintain state across infrastructure failures — e.g., webhook handlers that drop events, flaky cron jobs, background jobs that fail mid-execution, or workflows that need to resume where they left off. Covers Inngest function configuration, triggers (events, cron, invoke), step execution and memoization, idempotency, cancellation, error handling, retries, logging, and observability.
inngestinngest-steps
Use when implementing delays that must survive process restarts (e.g., 24-hour cart abandonment, scheduled follow-ups), waiting for human approval or external events with timeouts (review gates, webhook callbacks, async API completion), polling external services without losing state on crashes, calling other functions and awaiting their results, memoizing expensive operations so they don't re-run on retry, or running async work in parallel inside a workflow. Covers Inngest step methods: step.run, step.sleep, step.waitForEvent, step.waitForSignal, step.sendEvent, step.invoke, step.ai, plus patterns for loops and parallel execution.
inngest
plantuml-skill
Use when user requests diagrams, flowcharts, sequence diagrams, class diagrams, component diagrams, ER diagrams, architecture charts, or visualizations — including generating a diagram from existing source code, or rendering/extracting PlantUML embedded in a Markdown file to images (e.g. preparing docs for Confluence/Notion). Also use proactively when explaining systems with 3+ components, APIs, data flows, or class hierarchies. Generates .puml files and exports to PNG/SVG via Kroki API (no local install required).
agents365-ai
kling-3-0
>
agentspace-soface-swap
通过 `runcomfy` CLI 在 RunComfy 上将人脸/角色换入视频或图像。路由涵盖社区 Wan 2-2 Animate（音频驱动的角色动画 + 身份替换）、GPT Image 2 Edit（基于参考合成的单次精确换脸）、Nano Banana Edit（批量身份保持换脸）、Flux Kontext（单参考高保真局部人脸编辑）和 Kling 2-6 Motion Control Pro（将一段表演的动作迁移到目标角色上）。根据用户的实际意图选择正确的模型——单张图片 vs 视频、完整角色 vs 仅换脸、对话场景 vs 无声动作。触发词包括“换脸”、“交换人脸”、“deepfake”、“人脸替换”、“角色替换”、“换头”、“把 X 的脸放到 Y 上”、“让这个视频的主角变成 X”、“替换这个视频中的演员”、“交换照片中的角色”、“deepfake 视频”、“ReActor 替代品”，或任何明确要求将一个身份替换为另一个身份的请求。
agentspace-sovideo-inpainting
>
agentspace-soimage-inpainting
>
agentspace-soruncomfy-cli
从命令行在RunComfy上运行任何模型。`runcomfy` CLI是一个二进制文件、一次认证、数百个模型端点——图像生成、图像编辑、视频生成、图像转视频、唇形同步、换脸、视频编辑、修复、扩展、ControlNet、重打光、放大、LoRA训练等。提交请求、轮询状态、下载输出。本技能教代理如何安装、认证、发现模型模式、调用模型、流式/轮询/不等待、以JSON输出模式编写脚本以及处理错误。触发词包括"runcomfy cli"、"install runcomfy"、"runcomfy login"、"runcomfy run"、"runcomfy whoami"、"runcomfy api"，或任何明确要求从脚本或终端调用RunComfy模型的请求。同级技能（ai-image-generation、ai-video-generation、image-edit、video-edit、face-swap、lipsync、image-to-video、image-inpainting、image-outpainting、video-extend、controlnet-pose、relight）都通过此CLI调度。
agentspace-sovideo-extend
通过 `runcomfy` CLI 在 RunComfy 上扩展或延续现有视频片段。路由到 Google Veo 3-1 的 `extend-video` 和 `fast/extend-video` 端点——选择源视频并添加描述后续内容的提示词，模型会生成一个延续原视频的片段，保持运动、光照和主体一致性。当用户有一个较短的 Veo 片段并希望延长，或希望从单个种子片段逐镜头构建叙事链时使用。触发词包括“extend video”、“continue video”、“longer video”、“video extend”、“make this clip longer”、“Veo extend”、“chain video shots”、“video continuation”，或任何明确要求对现有视频添加更多帧的请求。
agentspace-soimage-outpainting
>
agentspace-solipsync
通过 `runcomfy` CLI 将面部与特定音频轨道进行唇形同步。路由覆盖 ByteDance OmniHuman（基于肖像+音频的音频驱动全身虚拟形象）、Sync Labs sync v2 / Pro（最先进的嘴部同步到视频）、Kling lipsync（音频转视频和文本转视频，带同步语音）以及 Creatify lipsync。该技能根据用户的实际意图选择正确的端点——静态肖像+音频（虚拟形象风格）、源视频+音频（现有素材上的嘴部替换）或从脚本生成并同步。触发词包括“lip sync”、“lipsync”、“make this video speak”、“match audio to mouth”、“dub video”、“sync lips to voice”、“Sync Labs”、“voiceover sync”或任何明确要求用音频轨道驱动面部嘴部的请求。
agentspace-sorelight
通过 `runcomfy` CLI 在 RunComfy 上重新打光静态图像——改变光照设置、色温、方向或氛围。路由到 Qwen Edit 2509 专用的 `relight` LoRA 端点进行针对性重打光，当文字描述足够时回退到保留身份的编辑端点（Nano Banana 2 Edit、GPT Image 2 Edit、FLUX Kontext Pro）。适用于产品重打光（影棚柔光箱→窗光）、人像氛围转换（阴天→黄金时刻）或色温调整。触发词包括“重打光”、“重新打光”、“改变光照”、“变成黄金时刻”、“影棚光”、“轮廓光”、“蓝色时刻”、“柔光窗光”、“改变光方向”、“色温”或任何明确要求改变静态图像光照的表述。
agentspace-so
chatgpt-codex-onboarding
Connect a ChatGPT or Codex subscription via OAuth device-code login. Use when the user wants to sign in with their ChatGPT Plus, Pro, or Team account (e.g. "use my Codex subscription", "log in with ChatGPT").
starchild-ai-agentimage-edit
Image editing and enhancement of an existing image. Covers background replacement, super-resolution upscaling, old photo restoration, colorization, person removal, portrait retouching (skin smoothing, blemish removal), slimming, color grading, artistic filters, image blending, outpainting, local editing, text rendering, multi-angle generation, before/after comparison, car recoloring, car wrap preview. Use when editing, enhancing, or transforming an existing image (e.g. remove background, upscale photo, restore old photo, retouch portrait, change car color, apply filter, extend image).
starchild-ai-agentimage-portrait
Identity-consistent portrait generation from a reference photo. Covers professional headshots, dating photos, style transfers, themed portraits, photo series, avatars, ID photos. Use when generating styled portraits from a reference photo (e.g. professional headshot, anime avatar, cyberpunk portrait, travel photo, dating profile photo, ID photo).
starchild-ai-agentnano-banana-edit
>
agentspace-sohappyhorse-1-0
使用 RunComfy 上的 HappyHorse 1.0 生成文本到视频。介绍 HappyHorse 1.0 的优势（Artificial Analysis Video Arena 排名第一，原生 1080p 并同步音频，多镜头角色一致性，支持 6 种语言提示），时长/宽高比/分辨率方案，以及何时改用 Wan 2.7 / Seedance 2 / LTX 2。通过本地 RunComfy CLI 调用 `runcomfy run happyhorse/happyhorse-1-0/text-to-video`。触发词包括“happyhorse”、“happy horse”、“happyhorse 1.0”、“happyhorse video”或任何明确要求使用此模型生成视频的请求。
agentspace-so