Agent Skill · 研究检索

ai-ugc（艾用户 GC）

ai-ugc 用于查找、检索和筛选相关信息，适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态，以及是否会触发联网、命令执行或文件读写。

ai-ugc 是研究检索类 Agent Skill，可用宿主为 OpenClaw，通过ClawHub获取，权限为接触敏感数据；许可证为 MIT-0，第三方扫描已通过，按安装量在研究检索类的 73,675 个 Skill 里排第 4,766。

总安装 29,911周安装 1,210GitHub Stars 2下载 9,390核验 2026-05-01

研究检索接触敏感数据 ClawHub未标认证来源可访问许可证明确审计通过 OpenClaw

安装说明

本站只整理中文说明，不托管安装包，也不代用户安装。

GitHub

许可证: MIT-0
版本: 0.0.4

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手，先确认命令和权限，再让它执行。

请帮我安装这个 Agent Skill：ai-ugc（艾用户 GC）
来源仓库：https://github.com/tfcbot/ai-ugc
安装命令：
openclaw skills install ai-ugc
安装前请先检查当前环境是否支持对应 CLI，并向我确认将要执行的命令、安装目录、联网范围和文件读写权限；确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill；本站只展示命令，不托管安装包，也不自动执行。

ClawHubOpenClaw

openclaw skills install ai-ugc

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题，通常会让助手先检索再总结。

用于查找、检索和筛选相关信息，适合在 OpenClaw 中根据关键词或任务场景快速定位候选结果。适用于内容生成前的素材调研、社交媒体趋势分析或创意灵感挖掘等研究场景。通过输入关键词、主题或来源线索，调用 RawUGC API 获取视频、图像或音乐等内容资源。

安装命令为 openclaw skills install ai-ugc，需确认权限范围及是否涉及联网或文件操作。建议结合原始 README 核验具体用法，并注意维护状态与版权合规要求。

上手步骤

4 步

1
确认自己的宿主工具
公开资料标注支持 OpenClaw；先确认自己用的是其中之一，再按对应工具的 Skills 目录放置。
2
拿到安装命令
可以直接复制的命令是 openclaw skills install ai-ugc；命令里的占位符换成自己的路径或密钥再执行。
3
安装后重启宿主
命令执行完要完全退出宿主工具再打开，让新装的 Skill 被重新扫描到，否则列表里可能看不到。
4
按最小权限试用
它需要接触敏感数据，先在测试目录或测试账号里跑，确认没问题再放到正式项目。

常见问题

6 条

ai-ugc 怎么安装？

在终端执行 openclaw skills install ai-ugc。装完完全退出宿主工具再打开，让它重新扫描本地 Skill。

ai-ugc 支持哪些 AI 工具？

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录，一般也能手动放进去，但要自己确认触发方式是否一致。

ai-ugc 需要什么权限？

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置，应进入高风险复核队列，默认不自动发布。

ai-ugc 是免费的吗？

来源标注的许可证是 MIT-0。本站不代收任何费用，也不托管安装包；能否商用、要不要署名，以许可证原文为准。

ai-ugc 用的人多吗？

公开数据是累计安装 29,911 次，近一周安装 1,210 次，下载 9,390 次，GitHub 2 星。这些数字来自第三方目录统计，只能当作热度参考，不代表质量或安全性。

ai-ugc 安全吗？

第三方扫描结果为通过，但这只覆盖已知的静态检查项，不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

专注 OpenClaw 一种宿主，触发方式和使用习惯不用来回切换。
许可证明确为 MIT-0，商用和分发边界好判断。
第三方扫描已通过，常见的高风险行为没有命中。

先想清楚的地方

需要接触敏感数据，装到常用环境前要先评估影响面。
第三方目录整理，版本更新和行为细节以来源仓库为准。

适不适合你

更适合

已经在用 OpenClaw，想直接扩展它能力的人。
需要处理研究检索类任务、又不想每次都手写提示词的人。
需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

环境里有大量密钥、凭证，不方便给第三方 Skill 读取的人。
需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类

Skill	总安装	GitHub 星	权限	许可证	扫描
find-skills 寻找技能 · Codex、Claude	31,199,813	16,486	操作浏览器	MIT	有提醒
self-improving-agent 自我改进 Agent · OpenClaw	16,464,002	3,396	接触敏感数据	MIT-0	通过
soultrace 灵魂追踪 · Codex、Claude	7,588,944	12	需要联网	未明确	有提醒
self-improving 自我完善 · OpenClaw	6,829,284	1,046	需要联网	MIT-0	通过
microsoft-foundry 微软代工厂 · Codex、Claude	6,614,144	764	调用外部服务	未明确	有提醒
ai-ugc 艾用户 GC · 本页	29,911	2	接触敏感数据	MIT-0	通过

按安装量取同分类前几名；数字来自第三方目录统计，只作热度参考，不代表质量结论。

SKILL.md 原文

name: rawugc-api
description: Call the RawUGC API to generate AI videos/images/music, manage content (personas, products, styles, characters), schedule social media posts, research TikTok content, and analyze viral videos. Use when the user wants to interact with any RawUGC API endpoint.
requires
env
compatibility: Requires RAWUGC_API_KEY (Bearer token for https://rawugc.com/api/v1). Obtain from RawUGC dashboard.
homepage: https://github.com/tfcbot/rawugc-skills
source: https://github.com/tfcbot/rawugc-skills

RawUGC API

Procedural knowledge for agents to call the RawUGC API. All requests require an API key from the RawUGC dashboard, passed via environment variable.

Authentication

Environment variable: Read the API key from RAWUGC_API_KEY. The key is created in the RawUGC dashboard and must be kept secret; do not hardcode or log it.
Header: Send on every request: Authorization: Bearer <value of RAWUGC_API_KEY>.
If RAWUGC_API_KEY is missing or empty, inform the user they must set it and obtain a key from the RawUGC dashboard.

Base URL

Production: https://rawugc.com/api/v1
All paths below are relative to this base.

API Versioning

RawUGC uses date-based API versioning. The current latest version is 2026-03-06.

RawUGC-Version request header: Override the version per-request (recommended).
API key pinned version: Set when creating the key in the dashboard.
Fallback: Latest version (2026-03-06) if neither is set.

Always send RawUGC-Version: 2026-03-06 in requests to ensure consistent behavior.

Video Generation

POST /videos/generate

Initiate video generation.

Request body (JSON):

Field	Type	Required	Description
`model`	string	Yes	`sora-2-text-to-video`, `sora-2-image-to-video`, `kling-2.6/motion-control`, `veo3`, `veo3_fast`
`prompt`	string	For text-to-video / veo3	Text description (1-5000 chars)
`imageUrls`	string[]	For image-to-video / kling	URLs, max 10. Veo3/veo3_fast accept up to 2 optional images.
`videoUrls`	string[]	For kling	URLs, max 1. Required for `kling-2.6/motion-control`
`aspectRatio`	string	No	Sora: `portrait`/`landscape`. Veo3: `16:9`/`9:16`/`Auto`
`nFrames`	string	No	`"10"` or `"15"` (Sora only)
`selectedCharacter`	string	No	Character username (e.g. `rawugc.mia`)
`characterOrientation`	string	No	`image` or `video` (kling only)
`mode`	string	No	`720p` or `1080p` (kling only)

Response (201): videoId, model, status, creditsUsed, newBalance, estimatedCompletionTime, createdAt.

GET /videos/:videoId

Get video status. Returns videoId, status, model, prompt, creditsUsed, url (when completed), createdAt, completedAt, failCode, failMessage, versions (edit history array).

GET /videos

List videos. Query: status, limit (1-100, default 50), page. Returns videos array + pagination.

POST /videos/captions

Add styled captions to a completed video. Costs 1 credit.

Field	Type	Required	Description
`videoId`	string	Yes	Video identifier (vid_xxx)
`language`	string	No	Language code (e.g. `en`). Defaults to auto-detect

Response (200): videoId, url, version, operation, creditsUsed.

POST /videos/overlay

Add text overlay to a completed video.

Field	Type	Required	Description
`videoId`	string	Yes	Video identifier (vid_xxx)
`text`	string	Yes	Overlay text (1-500 chars)
`position`	string	No	`top`, `center`, or `bottom`
`fontSize`	integer	No	8-200 pixels
`topBottomMargin`	integer	No	0-500 pixels
`strokeThickness`	number	No	0-10

Response (200): videoId, url, version, operation, creditsUsed.

Image Generation

POST /images/generate

Generate AI images using Nano Banana models. Async -- poll GET /images/:imageId.

Field	Type	Required	Description
`model`	string	Yes	`nano-banana-2` (text-to-image, 4 credits) or `google/nano-banana-edit` (image editing, 2 credits)
`prompt`	string	Yes	Text description or edit instruction (1-20000 chars)
`imageUrls`	string[]	For editing	Source images. Required for `google/nano-banana-edit`. Optional for `nano-banana-2` (reference images, max 14).
`aspectRatio`	string	No	For `nano-banana-2`: `1:1`, `16:9`, `9:16`, `auto`, etc.
`imageSize`	string	No	For `google/nano-banana-edit`: `1:1`, `16:9`, `9:16`, `auto`, etc.
`resolution`	string	No	For `nano-banana-2`: `1K`, `2K`, `4K`
`outputFormat`	string	No	`png`, `jpeg`, `jpg`
`googleSearch`	boolean	No	Use Google Web Search grounding (`nano-banana-2` only)

Response (201): imageId, model, status, creditsUsed, newBalance, estimatedCompletionTime, createdAt.

GET /images/:imageId

Get image status. Returns imageId, status, model, prompt, url (when completed), imageSize, resolution, outputFormat, creditsUsed, createdAt, completedAt, failCode, failMessage.

GET /images

List images. Query: status, limit (1-100, default 20), page. Returns images array + pagination.

Music Generation

POST /music/generate

Generate AI music using Suno models. 3 credits per generation. Async -- poll GET /music/:musicId.

Field	Type	Required	Description
`prompt`	string	Yes	Music description (1-2000 chars)
`model`	string	No	`V3_5`, `V4`, `V4_5`, `V4_5PLUS`, `V4_5ALL`, `V5` (default: `V5`)
`instrumental`	boolean	No	Instrumental only, no vocals (default: true)
`title`	string	No	Track title (max 200 chars). Enables custom mode with `style`.
`style`	string	No	Style descriptor (max 500 chars, e.g. `lo-fi hip hop`)

Response (201): musicId, model, status, creditsUsed, newBalance, estimatedCompletionTime, createdAt.

GET /music/:musicId

Get music status. Returns musicId, status, model, prompt, audioUrl (when completed), albumArtUrl, duration, title, creditsUsed, createdAt, completedAt, failCode, failMessage.

GET /music

List music tracks. Query: status, limit (1-100, default 20), page. Returns tracks array + pagination.

Upload

POST /upload

Upload a video or image file. Returns a URL for use in generation requests (imageUrls, videoUrls) or analyze-video. Max 100MB.

Request: multipart/form-data with file field. Accepted types: video/mp4, video/quicktime, video/webm, image/png, image/jpeg, image/webp.

Response (200): url, contentType, size.

Characters

GET /characters

List all available AI characters (built-in + custom). Returns characters array, count, adminCount, userCount.

GET /characters/:characterId

Get a character by ID. Returns _id, username, displayName, description, videoPreviewUrl, type (admin/user), isActive, createdAt, updatedAt.

Personas (CRUD)

Personas define target audiences for content plan generation.

GET /personas -- List all. Returns personas array + count.
POST /personas -- Create. Body: name (required, max 200), description (required, max 5000). Returns id.
GET /personas/:personaId -- Get one.
PATCH /personas/:personaId -- Update. Body: name, description (both optional).
DELETE /personas/:personaId -- Delete.

PersonaResponse: _id, organizationId, name, description, createdAt, updatedAt.

Messaging (CRUD)

Brand/positioning messaging templates.

GET /messaging -- List all. Returns messages array + count.
POST /messaging -- Create. Body: name (required, max 200), body (required, max 5000). Returns id.
GET /messaging/:messageId -- Get one.
PATCH /messaging/:messageId -- Update. Body: name, body (both optional).
DELETE /messaging/:messageId -- Delete.

MessagingResponse: _id, organizationId, name, body, createdAt, updatedAt.

Products (CRUD)

Products for video generation.

GET /products -- List all. Returns products array + count.
POST /products -- Create. Body: name (required, max 200), photos (required, URL array), description (max 1000), messaging (max 5000). Returns id.
GET /products/:productId -- Get one.
PATCH /products/:productId -- Update. Body: name, description, photos, messaging (all optional).
DELETE /products/:productId -- Delete.

ProductResponse: _id, name, description, photos, messaging, createdAt, updatedAt.

Styles (CRUD)

Video/image creative styles with optional prompt templates.

GET /styles -- List all (built-in + custom). Query: type (video/image). Returns styles array + count.
POST /styles -- Create. Body: name (required, max 200), description (max 1000), type (video/image), aspectRatio (portrait/landscape/square), promptTemplate (max 5000, supports {productName}, {messaging}, {character} placeholders). Returns id.
GET /styles/:styleId -- Get one.
PATCH /styles/:styleId -- Update. All fields optional.
DELETE /styles/:styleId -- Delete.

StyleResponse: _id, name, description, type, aspectRatio, styleId, promptTemplate, isAdmin, isStandard.

Social Scheduling

GET /social/accounts

List connected social accounts (max 3 per org). Returns accounts array + count. Each account: accountId, platform (tiktok/instagram/youtube), username, displayName, profilePicture, isActive.

POST /social/accounts

Sync connected accounts from the scheduling provider. Returns { success: boolean }.

DELETE /social/accounts/:accountId

Disconnect a social account. Returns { success: boolean }.

POST /social/posts

Schedule, draft, or immediately publish a video to social media.

Field	Type	Required	Description
`videoUrl`	string	Yes	URL of video to post
`accountIds`	string[]	Yes	Target account IDs
`mode`	string	Yes	`schedule`, `draft`, or `now`
`scheduledFor`	integer	For schedule	Unix timestamp (ms)
`timezone`	string	No	IANA timezone (default: `UTC`)
`content`	string	No	Caption (max 2200 chars)
`videoId`	string	No	RawUGC video ID to link
`publishToInbox`	boolean	No	Send to TikTok Creator Inbox
`tiktokPrivacyLevel`	string	No	`SELF_ONLY`, `PUBLIC_TO_EVERYONE`, `MUTUAL_FOLLOW_FRIENDS`, `FOLLOWER_OF_CREATOR`
`tiktokAllowComment`	boolean	No	Allow TikTok comments
`tiktokAllowDuet`	boolean	No	Allow TikTok duets
`tiktokAllowStitch`	boolean	No	Allow TikTok stitches
`tiktokCommercialContentType`	string	No	`none`, `brand_organic`, `brand_content`

Response (201): SocialPost object.

GET /social/posts

List posts. Query: fromDate (ms), toDate (ms), includeDrafts (boolean). Returns posts array + count.

GET /social/posts/:postId

Get a post.

PATCH /social/posts/:postId

Update a post. Body: content, scheduledFor, timezone, accountIds (at least one field required).

DELETE /social/posts/:postId

Delete a post. Returns { success: boolean }.

POST /social/posts/:postId/reschedule

Reschedule a post. Body: scheduledFor (required, ms), timezone.

POST /social/posts/:postId/publish

Immediately publish a draft post.

SocialPost: postId, platforms, status (draft/scheduled/published/failed), scheduledFor, timezone, content, videoUrl, createdAt, publishedAt.

Viral Library

GET /viral-library/videos/:videoId

Get a viral library video with full AI analysis (hooks, keyframes, performance insights). Returns ViralLibraryVideo.

GET /viral-library/search

Semantic search across analyzed videos. Query: q (required, natural language), limit (1-50, default 20). Returns results (array of { video, score }), query, total.

Research

POST /scrape-tiktok

Scrape TikTok videos. Costs 3 credits.

Field	Type	Required	Description
`query`	string	Yes	Search keyword, hashtag, or query (max 500)
`mode`	string	No	`keyword`, `hashtag`, `search` (default: `keyword`)
`limit`	integer	No	1-10 (default: 10)

Response (200): scrapeId (use with content-plans), count, videos (array with id, url, author, description, stats, duration, hashtags, thumbnail, videoUrl).

POST /content-plans

Generate a content plan from scraped videos. Costs 3 credits.

Field	Type	Required	Description
`scrapeId`	string	Yes	From scrape-tiktok response
`brief`	string	Yes	Content plan goals (max 5000)

Response (200): planId, scrapeId, brief, topWins, gapsToTest, blueprints (array with category, strategy, evidence, contentIdeas).

GET /content-plans

List all content plans. Returns plans array + count.

POST /analyze-video

Analyze any video URL (social links or direct URLs). Costs 1 credit. Max 150MB.

Field	Type	Required	Description
`videoUrl`	string	Yes	Video URL to analyze
`prompt`	string	No	Custom analysis prompt (max 5000)

Response (200): summary, hook, keyframes (array with timestamp, type, description, visual, audio, text), durationSeconds, tags, whyItPerformed, attributesToCopy, hooksToTest.

Errors

All error responses use RFC 7807 Problem Details (JSON): type, title, status, detail, instance, errors.

Status	Meaning
400	Validation error. Surface `detail` and `errors` to user.
401	Auth error. Check `RAWUGC_API_KEY`.
402	Insufficient credits. Add credits in dashboard.
403	Insufficient scope. API key lacks permissions.
404	Resource not found.
429	Rate limit exceeded. Check `X-RateLimit-Reset` header.
500	Server error. Retry or contact support.

Rate Limits

API Key: 10 req/min. Session: 20 req/min.
Headers: X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset (unix timestamp).

Workflow: Generate then poll

Generate: POST to the generation endpoint. Note the returned ID (videoId/imageId/musicId).
Poll: GET the status endpoint periodically (10-30s). Use exponential backoff.
Finish: When status === 'completed', use the result URL. When failed, surface error to user.
Edit (video only): POST to /videos/captions or /videos/overlay.

For full request/response shapes, see reference.md.

适合场景

文本生成图片

图片风格化

产品图和创意图

需要 FLUX 模型时

能力概览

能力 1

调用 FLUX 图像模型

能力 2

支持文本生图和图像改写

能力 3

覆盖 LoRA 或风格适配

能力 4

适合创意视觉生成

平台分布

OpenClaw

92.28%

按下载量换算8,665

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置，应进入高风险复核队列，默认不自动发布。

安装前确认

本站仅展示第三方公开信息，不托管安装包，不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源，建议同时核对源仓库。

继续浏览同类 Skills

更多研究检索 Skills（73,675）返回 Skills 目录

上一个 Skill

ai-twitter-daily（艾推特日报）

danetteceola/ai-twitter-daily

ai-twitter-daily 用于查找、检索和筛选相关信息，适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态，以及是否会触发联网、命令执行或文件读写。

下一个 Skill

ai-ugc-ads（AI UGC 广告）

chadboyda/agent-gtm-skills

ai-ugc-ads 用于查找、检索和筛选相关信息，适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态，以及是否会触发联网、命令执行或文件读写。