
Xyq Nest Skill
Drive 小云雀 NestAgent sessions from your repo to generate, edit, and download AI images and videos via Python CLIs.
Overview
xyq-nest-skill is an agent skill most often used in Build (also Launch distribution, Grow content) that orchestrates 小云雀 NestAgent for AI image and video generation, editing, and download.
Install
npx skills add https://github.com/pippit-dev/pippit-skills --skill xyq-nest-skillWhat is this skill?
- Four scripted flows: create/send message, poll thread progress, upload assets, parallel download of outputs
- Text-to-image, text-to-video, image-to-video, video continuation on the Nest platform
- Edit modes: local edits, element swap, camera moves, style transfer
- Complex pipelines: one-line short drama, TVC/replica, music MV, product showcase
- Python 3.6+ MIT skill with thread_id and run_id returned from submit_run.py
- 4 documented capability rows: session/messaging, progress query, upload, download
- Python 3.6+ requirement
Adoption & trust: 2.8k installs on skills.sh; 2/3 security scanners passed (skills.sh audits).
What problem does it solve?
You need programmatic control over AI video and image creation without manually operating a separate creative console.
Who is it for?
Indie makers automating promo clips, MVs, or product films through Claude/Cursor-driven Python scripts.
Skip if: Teams that only need static SVG/UI mocks with no generative video API or Nest account access.
When should I use this skill?
When you need to create, edit, or download AI images or videos through 小云雀 NestAgent from the terminal.
What do I get? / Deliverables
You get thread and run IDs, polled artifacts, and local files ready for edits, campaigns, or product embeds.
- thread_id and run_id from submissions
- Polled message sequences and artifact metadata
- Downloaded image/video files on disk
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Primary shelf is Build integrations because the skill is an API/script bridge to an external creative agent, not a standalone editor UI. Integrations subphase covers thread runs, asset uploads, and polling—classic third-party service wiring for solo builders.
Where it fits
Submit a text-to-video prompt and poll until run artifacts land in your repo output folder.
Regenerate a TVC variant from an uploaded reference asset ID before a paid campaign goes live.
Continue an existing thread to produce a second story video for a lifecycle email series.
How it compares
Python integration skill to a hosted generative studio—not a local Stable Diffusion CLI or generic FFmpeg wrapper.
Common Questions / FAQ
Who is xyq-nest-skill for?
Solo builders and small content teams who want agent-driven sessions against 小云雀 NestAgent for images and video.
When should I use xyq-nest-skill?
In Build when wiring creative APIs; in Launch when batching ad or social assets; in Grow when refreshing lifecycle creatives from the same thread.
Is xyq-nest-skill safe to install?
It runs local Python against a remote service—review the Security Audits panel on this page and treat API tokens and downloaded media as sensitive.
SKILL.md
READMESKILL.md - Xyq Nest Skill
*.pyc # 🎬 xyq-nest-skill [](../../LICENSE) [](https://python.org) ## 🚀 简介 `xyq-nest-skill` 通过接入小云雀NestAgent的综合创作能力,实现AI图片/视频的生成、编辑、风格转换等功能。 ## ✨ 功能特性 | 功能 | 说明 | |------|----------------------------------| | 📨 创建会话 / 发送消息 | 向小云雀发送自然语言指令,生成图片或视频 | | 🔍 查询会话进展 | 增量拉取会话消息,轮询创作进度和产物结果 | | 📤 上传文件 | 上传图片/视频到小云雀资产库,获取 `资产ID` 用于编辑和参考 | | 📥 下载结果 | 批量下载生成的图片/视频到本地,支持并行下载 | **🎨 小云雀平台能力覆盖:** - **生成**:文生图、文生视频、图生视频、视频续写 - **编辑**:局部修改、元素替换、镜头调整、风格迁移 - **复杂创作**:一句话生成短剧(剧本→分镜→成片)、复刻视频/TVC/宣传片、音乐 MV 生成、产品展示片制作 ### 📨 创建会话 / 发送消息 ```bash # 创建新会话 python3 skills/xyq-nest-skill/scripts/submit_run.py --message "生一个动漫视频" # 向已有会话发送消息 python3 skills/xyq-nest-skill/scripts/submit_run.py --message "再生成一个故事视频" --thread-id THREAD_ID # 携带参考文件发送 python3 skills/xyq-nest-skill/scripts/submit_run.py --message "参考这个视频做修改" --asset-ids asset_id1 asset_id2 ``` **参数:** | 参数 | 必填 | 说明 | |------|------|------| | `--message` | 是 | 创作指令内容 | | `--thread-id` | 否 | 已有会话 ID,不传则创建新会话 | | `--asset-ids` | 否 | 资产 ID 列表,支持多个 | **返回:** ```json { "thread_id": "90f05e0c-...", "run_id": "abc123-..." } ``` ### 🔍 查询会话进展 ```bash python3 skills/xyq-nest-skill/scripts/get_thread.py --thread-id THREAD_ID --run-id RUN_ID --after-seq 0 ``` **参数:** | 参数 | 必填 | 说明 | |------|------|------| | `--thread-id` | 是 | 会话 ID | | `--run-id` | 否 | 运行 ID | | `--after-seq` | 否 | 增量拉取起始序号,默认 0 | **返回:** ```json { "messages": [ {"id": "1", "role": "user", "content": "生一个动漫视频"}, {"id": "2", "role": "assistant", "content": [{"type": "...", "subtype": "...", "data": {...}}]}, {"id": "3", "role": "assistant", "content": [{"type": "...", "data": {"url": "https://..."}}]} ] } ``` ### 📤 上传文件 ```bash # 上传图片 python3 skills/xyq-nest-skill/scripts/upload_file.py /path/to/image.png # 上传视频 python3 skills/xyq-nest-skill/scripts/upload_file.py /path/to/video.mp4 ``` 仅支持 `image/*` 和 `video/*` 类型,单文件大小限制 200MB。 **返回:** ```json { "asset_id": "asset_xxx" } ``` ### 📥 下载结果 ```bash python3 skills/xyq-nest-skill/scripts/download_results.py \ --urls URL1 URL2 URL3 \ --output-dir ./xyq_output \ --prefix "storyboard" \ --workers 5 ``` **参数:** | 参数 | 必填 | 说明 | |------|------|------| | `--urls` | 是 | 要下载的 URL 列表 | | `--output-dir` | 否 | 输出目录,默认 `./xyq_output` | | `--prefix` | 否 | 文件名前缀,如 `storyboard` → `storyboard_01.png` | | `--workers` | 否 | 并行下载线程数,默认 5 | **返回:** ```json { "output_dir": "./xyq_output", "downloaded": ["./xyq_output/storyboard_01.png", "..."], "total": 3 } ``` ## 🎯 典型示例 ### 🎬 文生视频(最常见) ``` 1. submit_run.py --message "生成一个赛博朋克风格的城市夜景视频" → 拿到 thread_id、run_id 2. 每 10 秒轮询: get_thread.py --thread-id THREAD_ID --run-id RUN_ID --after-seq SEQUENCE → ⏳ 进行中:展示创作信息,继续轮询 → ❓ 意图确认:展示问题,等用户回复后用同一 thread_id 重新提交 → ✅ 完成:提取产物 URL,下载并展示结果 3. download_results.py --urls URL1 URL2 --output-dir ./output --prefix "cyberpunk" ``` ### ✂️ 编辑已有视频 ``` 1. upload_file.py /path/to/video.mp4 → 拿到 asset_id 2. submit_run.py --message "把背景换成星空" --asset-ids asset_id 3. 同场景 1 的轮询和下载流程 ``` ### 🖼️ 多参考图生成 ``` 1. upload_file.py /path/to/ref1.png → asset_id1 2. upload_file.py /path/to/ref2.png → asset_id2 3. upload_file.py /path/to/ref3.mp4 → asset_id3 (多文件可并行上传) 4. submit_run.py --message "根据参考图和视频生成科普故事视频" --asset-ids asset_id1 asset_id2 asset_id3 5. 同场景 1 的轮询和下载流程 ``` ### 💬 在已有会话中追加需求 ``` 1. submit_run.py --message "把刚才的视频加个片头" --thread-id EXISTING_THREAD_ID → 新的 run_id 2. 同场景 1 的轮询和下载流程 ``` ### ⏱️ 轮询策略 - **间隔**:每 10 秒查询一次 - **增量拉取**:首次 `--after-seq 0`,后续根据已获取消息数计算新 seq - **超时**:连续轮询 48 小时无结果则停止 - **错误重试**:单次失败可重试 1 次,连续 3 次失败则停止 --- name: xyq-nest-skill description: 通过小云雀的 AI 能力进行综合创作,支持生成和编辑图片/视频。覆盖场景包括:生成(文生图、文生视频、图生视频、做动画、画一个xxx、来段xxx)、编辑修改(把xxx换成yyy、去掉xxx、加上xxx、改成xxx、调整xxx、局部修改、改镜头)、风格转换(风格迁移、转绘、换风格)、视频续写延长、复刻视频/TVC/宣传片、短剧/短漫剧生成、音