LinkFoxAI

v1.3.0

调用 LinkFox AI 开放平台 AI 作图/视频 API(换模特、抠图、扩图、场景裂变、智能修图、套图、姿势裂变、带货口播、图转视频等)。适用场景:(1) AI 任务创建与结果轮询,(2) 通过 api-call 调用任意开放平台接口,(3) 素材连通性测试。

0· 212·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for linkfox-ai/linkfox-ai.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "LinkFoxAI" (linkfox-ai/linkfox-ai) from ClawHub.
Skill page: https://clawhub.ai/linkfox-ai/linkfox-ai
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Required env vars: LINKFOXAI_API_KEY
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install linkfox-ai

ClawHub CLI

Package manager switcher

npx clawhub@latest install linkfox-ai
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description match the provided files and the script implements the advertised image/video creation, upload, polling and generic api-call functionality. Required env var LINKFOXAI_API_KEY is appropriate for the described API access.
Instruction Scope
SKILL.md confines runtime actions to calling LinkFox endpoints and optionally reading local image files for base64 upload. It recommends using sessions_spawn for long-running tasks. Note: the script supports reading additional environment variables (LINKFOXAI_APP_ID, LINKFOXAI_APP_SECRET, LINKFOXAI_BASE_URL, LINKFOXAI_IMAGE_EDIT_PATH) that are not listed in the skill metadata; this is a benign capability but worth noting.
Install Mechanism
No install spec; instruction-only plus a bundled Python script. Nothing is downloaded from external URLs during install.
Credentials
Declared required env var is LINKFOXAI_API_KEY which is appropriate. The script also accepts LINKFOXAI_APP_ID and LINKFOXAI_APP_SECRET (alternative auth) and optional LINKFOXAI_BASE_URL / LINKFOXAI_IMAGE_EDIT_PATH overrides; these additional env vars are reasonable but were not declared in the registry metadata.
Persistence & Privilege
Skill is not always-enabled and does not request elevated platform privileges. It does not modify other skills or system-wide configs; it only performs network calls when invoked.
Assessment
This skill appears to do exactly what it says: call LinkFox AI image/video APIs, upload images, and poll for results. Before installing: (1) Only provide an API key you trust (LINKFOXAI_API_KEY); the script also supports APP_ID/APP_SECRET if you prefer that auth method. (2) Be aware the tool can read local image files when you run upload-base64 — don’t point it at sensitive files. (3) The skill can be pointed at custom base URLs or a proxy (the docs mention a proxy host and an IP to whitelist); only use those if you trust the proxy. (4) The skill makes outbound network requests to the LinkFox endpoints; verify those endpoints are acceptable for your environment. If you need tighter guarantees, ask the publisher for a signed provenance or run the bundled script in a sandbox first.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

🦊 Clawdis
EnvLINKFOXAI_API_KEY
latestvk978k4y3vpnzk4n71tzzwyb6z184hexy
212downloads
0stars
3versions
Updated 2w ago
v1.3.0
MIT-0

LinkFoxAI - LinkFox AI 开放平台 Skill

LinkFoxAI 让 OpenClaw 调用 LinkFox AI 开放平台 AI 作图/视频能力:换模特、自动抠图、场景裂变、智能扩图、高清放大、消除笔、智能修图、商品套图、服装套图、姿势裂变、带货口播、图转视频等。

配置

  1. 获取凭证:注册 紫鸟开放平台,认证开发者并创建 LinkFox 卖家应用,然后在 linkfox.com/team/api-guide 将应用绑定到团队,获取 API Key。
  2. 环境变量LINKFOXAI_API_KEY(单一 API Key,配置一个 Key 即可调用全部接口)
  3. 自定义 Base URL(可选):默认 https://sbappstoreapi.ziniao.com/openapi-router。若无固定 IP,设置 LINKFOXAI_BASE_URL=https://sbappstoreapi-proxy.linkfox.com 并配置白名单。

必须使用 sessions_spawn 调度作图任务

AI 作图任务需要排队和生成时间(通常 10 秒 ~ 5 分钟,极端情况下可能因服务端超时或队列积压导致单个任务耗时超过 30 分钟)。务必通过 sessions_spawn 将作图任务派发到子会话,让主会话保持响应。脚本的 --wait 参数会自动轮询直到出结果。

超时处理策略(AI 自行判断)

当作图任务超时(--timeout 耗尽仍未返回结果)时,AI 应根据具体情况自主选择以下策略,无需等待用户指令:

  1. 延长轮询时间:将 --timeout 增大到 1800(30 分钟),--interval 建议 10~30 秒,适用于重要任务或用户明确愿意等待的场景。
  2. 告知用户并继续后台等待:向用户说明任务仍在生成中,预估剩余时间,并在后台子会话中继续轮询;完成后主动汇报结果。
  3. 任务分拆重试:若任务参数含多图(如 --output-num 4),可拆分为多个单图任务并行提交,降低单个任务超时概率。
  4. 放弃并报错:仅当用户明确要求快速返回、或超时时间已大幅超出合理范围(> 40 分钟)时,才终止轮询并告知用户稍后手动查询任务 ID。

默认建议:将 sessions_spawnrunTimeoutSeconds 设为 1800--timeout 设为 1700,以覆盖绝大多数超时场景。仅在明确知道该类型任务通常很快时才缩小超时值。

标准调度模板

sessions_spawn:
  task: |
    Run: <skill>/scripts/linkfoxai.py change-model --wait --timeout 1700 \
         --image-url "<模特原图>" --head-url "<头部图>" --output-num 4
    解析输出 JSON,提取 resultList 中的图片 URL 汇报给用户。
    如果 status=4(失败),报告 errorMsg。
    如果超时仍未完成,告知用户任务 ID 以便后续手动查询。
  label: "LinkFoxAI: AI 换模特"
  mode: "run"
  runTimeoutSeconds: 1800
  cleanup: "keep"

并行调度多个独立任务

# 子任务 1
sessions_spawn:
  task: "Run: <skill>/scripts/linkfoxai.py cutout --wait --timeout 1700 --image-url '<url>' --sub-type 1 ..."
  label: "LinkFoxAI: 自动抠图"
  mode: "run"
  runTimeoutSeconds: 1800

# 子任务 2
sessions_spawn:
  task: "Run: <skill>/scripts/linkfoxai.py scene-fission --wait --timeout 1700 --image-url '<url>' ..."
  label: "LinkFoxAI: 场景裂变"
  mode: "run"
  runTimeoutSeconds: 1800

通用接口调用(api-call)

对于脚本没有专属命令的接口(如 AI 穿衣、商品替换、饰品文字等),使用 api-call 传入路径和 JSON body:

sessions_spawn:
  task: |
    Run: <skill>/scripts/linkfoxai.py api-call \
         --path /linkfox-ai/image/v2/make/fittingRoom \
         --body '{"upperOriginUrl":"<上衣图>","modelImageUrl":"<模特图>","outputNum":2}'
    获取返回的 data.id,然后运行:
    <skill>/scripts/linkfoxai.py poll --id <task_id> --timeout 1700
    解析结果汇报给用户。如果超时仍未完成,告知用户任务 ID 以便后续手动查询。
  label: "LinkFoxAI: AI 穿衣"
  mode: "run"
  runTimeoutSeconds: 1800

脚本命令一览

基础命令

命令说明
material-list --type <1-10>作图素材列表(可用于连通性测试)
upload-base64 --file <路径>base64 上传图片,返回 viewUrl
make-info --id <任务ID>查询作图结果(单次)
poll --id <任务ID> [--interval 3] [--timeout 300]轮询作图结果直到成功/失败
refresh --id <图片ID> [--format jpg]刷新结果图片地址(注意是图片 ID,非任务 ID)
api-call --path <路径> --body '<JSON>'通用 API 调用,可调任意开放平台接口

作图快捷命令(均支持 --wait

命令说明
change-model --image-url <url> --head-url <url> [--scene-url <url>] [--output-num 1-4]AI 换模特
cutout --image-url <url> --sub-type <1/2/3/9/12/13> [--cloth-class tops,coat,...]自动抠图
scene-fission --image-url <url> [--strength 0.5] [--prompt "描述"] [--provider SCENE_FISSION_REALISTIC]场景裂变
expand-image --image-url <url> --width 1024 --height 1024 [--prompt "描述"]智能扩图
super-resolution --image-url <url> --magnification 2 [--enhance]图片高清放大
image-edit --image-url <url> --prompt "描述" [--provider BANANA_PRO/WAN2_7/SEEDREAM5] [--template 白底图]智能修图
erase --image-url <url> --mask-url <url>消除笔
sales-video --prompt "口播文案" --video-type WAN|SEED [--image-list <url1> <url2>] [--video-time 10] [--aspect-ratio 9:16] [--is-pro]带货口播(WAN: 10/15秒, SEED: 5/10/15秒)
image-to-video --image-url <url> --video-type SEED [--prompt "描述"] [--video-time 10] [--is-pro] [--aspect-ratio 9:16]图转视频(KLING/V/WAN/SEED/HAILUO)

所有作图快捷命令均支持 --wait [--timeout 300] [--interval 3],提交任务后自动轮询直到完成。

其他能力(通过 api-call 调用)

以下能力通过 api-call --path <路径> --body '<JSON>' 调用,接口详情见 references/image-make.md

能力路径
AI 换模特 2.0/linkfox-ai/image/v2/make/changeModelFixed
模特换场景/linkfox-ai/image/v2/make/modelChangeScene
AI 穿衣-上下装/linkfox-ai/image/v2/make/fittingRoom
AI 穿衣-连体衣/linkfox-ai/image/v2/make/fittingRoomSuit
AI 穿戴/linkfox-ai/image/v2/make/intelligentWear
商品替换/linkfox-ai/image/v2/make/shopReplace
场景图生成/linkfox-ai/image/v2/make/aiDraw
相似图裂变/linkfox-ai/image/v2/make/imageToImage
精细抠图-创建/linkfox-ai/v2/process/result/interactCutout/create
精细抠图-结果/linkfox-ai/v2/process/result/interactCutout/info
图片翻译/linkfox-ai/image/v3/make/imageTransl
手部修复/linkfox-ai/image/v2/make/handRepair
局部重绘/linkfox-ai/image/v2/make/areaRepair
印花提取/linkfox-ai/image/v2/make/printExtract
手持商品/linkfox-ai/image/v2/make/handHeld
饰品文字/linkfox-ai/image/v2/make/linkedWord
商品精修/linkfox-ai/image/v2/make/productRepair
图片获取描述词-创建/linkfox-ai/v2/process/result/imageToPrompt/create
图片获取描述词-结果/linkfox-ai/v2/process/result/imageToPrompt/info
智能修图/linkfox-ai/image/v2/make/imageEditV2
智能修图-多图/linkfox-ai/image/v2/make/multiImageFusionV2
商品套图/linkfox-ai/image/v2/make/productMarketMaterialV3
服装套图/linkfox-ai/image/v2/make/wearCollectionV2
姿势裂变/linkfox-ai/image/v2/make/modelPoseFission
带货口播/linkfox-ai/image/v2/make/salesVideo
图转视频/linkfox-ai/image/v2/make/imageToVideo

连通性测试

使用作图素材接口验证凭证和网络是否正常(无需上传图片):

<skill>/scripts/linkfoxai.py material-list --type 1

或对 OpenClaw 说:「用 LinkFoxAI 拉取作图素材列表」。

错误处理

  • 错误码(如 ERR_FORBIDDEN、ERR_NOT_HAVE_TEAM_PRIVILEGE、TEAM_PERMISSION_EXPIRE)见 references/open-platform.md
  • 需先将应用绑定团队且团队已开通套餐。
  • 作图任务失败时 status=4,检查 errorCodeerrorMsg
  • 常见问题:参数校验失败(ERR_FILED_VALIDATE)、图片审核不通过(ERR_IMAGE_MAKE_AUDIT)、点数不足(ERR_IMAGE_MAKE_WILL_OUT)。
  • 若调用新增能力,需使用对应 V2/V3 路径,否则可能出现参数不兼容或能力不可用。

参考文档

  • references/open-platform.md — 接入流程、API 基础地址、错误码
  • references/image-make.md — 全部作图接口及带货口播接口的完整参数与注意事项

示例

  1. 连通性:「用 LinkFoxAI 拉取作图素材列表」→ 脚本执行 material-list --type 1
  2. AI 换模特:「用 LinkFoxAI 做一次 AI 换模特,原图 xxx,头部图 xxx,出 4 张」→ change-model --wait --output-num 4
  3. 自动抠图:「用 LinkFoxAI 把这张图抠图」→ cutout --wait --sub-type 1
  4. 智能修图:「用 LinkFoxAI 把这张图改成白底图」→ image-edit --wait --prompt "白底图" --template 白底图(高质量:加 --provider BANANA_PRO
  5. 带货口播:「用 LinkFoxAI 生成一条 10 秒带货口播视频,竖版」→ sales-video --wait --video-type WAN --video-time 10 --aspect-ratio 9:16 --prompt "..."
  6. 带货口播(SEED):「用 LinkFoxAI 生成一条 SEED 15 秒口播视频」→ sales-video --wait --video-type SEED --video-time 15 --prompt "..."
  7. 图转视频:「用 LinkFoxAI 把这张图生成 SEED 视频」→ image-to-video --wait --image-url <url> --video-type SEED --video-time 10
  8. 通用调用:「用 LinkFoxAI 做 AI 穿衣」→ api-call --path /linkfox-ai/image/v2/make/fittingRoom --body '{...}' + poll --id <id>

Comments

Loading comments...