← Back to Skills Marketplace
priesttomb

skill-test-demo

by PriestTmob · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
39
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install my-skill-test
Description
用于调用 ZingAPI 的“创意列表查询”接口。适用于按行业、国家/地区、渠道、关键词、素材类型、AI 标签、时间范围、排序方式等条件检索广告创意、广告素材、创意投放记录、广告主创意和近期新增创意。
README (SKILL.md)

\r \r

ZingAPI 创意列表查询\r

\r 本 skill 用于调用 ZingAPI 的创意列表查询接口,并将自然语言检索需求转换为稳定的接口请求。\r \r 接口地址:\r \r POST https://openapi.dataideaglobal.com/zingapi/v1/creative/list/{customer_name}\r \r 固定请求头:\r \r x-zf-action: creative-list\r \r

凭证要求\r

\r 调用接口前必须配置以下环境变量:\r \r

  • ZINGAPI_ACCESS_KEY_ID\r
  • ZINGAPI_ACCESS_KEY_SECRET\r
  • ZINGAPI_CUSTOMER_NAME\r \r 如任一变量缺失,应先提示用户完成配置。不得在回复、日志或错误信息中展示密钥明文。\r \r

使用方式\r

\r 所有接口请求均通过 scripts/creative_list.py 执行。常规查询优先使用脚本参数;需要完整控制入参时,可使用 JSON 请求体。\r \r 示例:查询最近 30 天美国 Facebook 渠道的游戏视频创意,并返回 AI 标签。\r \r

python scripts/creative_list.py `\r
  --app-type 游戏 `\r
  --geo 美国 `\r
  --platform Facebook `\r
  --creative-type 视频 `\r
  --seen-days 30 `\r
  --sort 最新 `\r
  --include-ai-tags\r
```\r
\r
示例:使用原始 JSON 请求体。\r
\r
```powershell\r
python scripts/creative_list.py --body '@request.json' --output summary\r
```\r
\r
示例:仅检查请求体与签名头,不发送真实请求。\r
\r
```powershell\r
python scripts/creative_list.py --app-type 游戏 --keyword puzzle --dry-run\r
```\r
\r
## 参数转换规范\r
\r
根据用户描述转换为脚本参数时,遵循以下规则:\r
\r
- 完整入参映射见 `references/input-mapping.md`。当用户描述命中该映射表中的字段,但脚本没有对应的专用命令行参数时,应生成 JSON 请求体,并使用 `python scripts/creative_list.py --body '@request.json'` 调用。\r
- 行业类型:`游戏`、`工具`、`电商` 分别对应 `app_type` 的 `1`、`2`、`3`。\r
- 素材类型:`图片`、`视频`、`轮播`、`HTML`、`试玩` 分别对应 `ads_type` 的 `1`、`2`、`3`、`4`、`7`。\r
- 国家/地区:中文国家名应转换为三位国家/地区编码,例如 `美国` 转为 `USA`,`日本` 转为 `JPN`。\r
- 排序方式:`最新`、`展现`、`热度`、`最后发现`、`投放天数`、`点赞`、`评论`、`分享` 分别转换为对应的 `sort_field`。\r
- 去重方式:`广告去重`、`素材去重`、`广告主去重` 分别对应 `duplicate_removal` 的 `0`、`1`、`2`。\r
- 页码默认值为 `page=1`;每页数量默认值为 `page_size=20`,最大值为 `20`。\r
- 用户未指定时间范围时,默认查询最近 7 天。\r
- 用户要求“返回 AI 标签”或“带 AI 标签”时,添加 `--include-ai-tags`。\r
\r
## 执行流程\r
\r
1. 识别用户的检索目标、行业、地区、渠道、素材类型、时间范围和排序方式。\r
2. 若关键条件缺失但可使用默认值,应直接采用默认值;仅在业务目标无法判断时再询问用户。\r
3. 使用 `scripts/creative_list.py` 发起请求。\r
4. 默认返回摘要结果;仅当用户明确要求完整响应时使用 `--output raw`。\r
5. 错误处理时保留 `trace_id`,便于接口问题排查。\r
6. 如后续需要查询创意详情,应保留结果中的 `ad_key`、`app_type` 和 `search_flag`。\r
\r
## 输出字段\r
\r
默认摘要结果按业务含义分组展示,包括:\r
\r
- `广告主`:广告主名称、ID 或域名、开发者、Logo、分类。\r
- `素材`:素材类型、预览图、资源列表、视频时长、素材 ID、NSFW 标记。\r
- `投放`:行业、渠道、国家地区、文案语言、首次投放时间、最后发现时间、投放天数。\r
- `指标`:人气总值、当周人气值总和、展现估值、热度值、关联广告数、互动数据。\r
- `文案`:标题、正文、Message、CTA。\r
- `链接`:落地页、原帖链接、结束卡片或关联视频。\r
- `游戏标签`:游戏核心赛道、游戏 IP、游戏玩法、游戏主题。\r
- `AI素材标签`:图片、视频及新版 AI 标签。
- `原始字段`:保留关键原始字段,便于二次查询和排查。

字段含义以 `references/response-fields.md` 为准。用户要求展示“人气值”时,应使用 `指标.人气总值` 或 `指标.当周人气值总和`;不得将 `指标.热度值` 作为人气值展示。
摘要结果会将渠道、国家/地区等字典编码转换为“名称 (编码)”格式。人气值、热度值、展现估值等较大数值会使用 `K`、`M`、`B` 缩写并保留一位小数;原始编码和原始数值保留在 `原始字段` 中。

当 `is_nsfw` 为 `true` 时,素材资源地址可能被隐藏或替换为默认图片。
\r
## 参考资料\r
\r
- 参数说明见 `references/parameters.md`。\r
- 完整入参映射见 `references/input-mapping.md`。\r
- 出参字段说明见 `references/response-fields.md`。\r
- 配置与调用示例见 `references/examples.md`。\r
Usage Guidance
Install only if you intend to let the agent query ZingAPI with your ZingAPI credentials. Keep the access key secret out of chat/log output, review any raw JSON request body before sending it, and use dry-run when you want to inspect the signed request without making the API call.
Capability Assessment
Purpose & Capability
The skill purpose, docs, agent prompt, and script all align around querying ZingAPI creative-list data and summarizing returned advertising creative fields.
Instruction Scope
It asks for ZingAPI environment credentials, can read a user-specified JSON body via @path, and can send signed requests; these behaviors are disclosed and tied to the API workflow, though the manifest does not provide a separate permission boundary.
Install Mechanism
No installer, dependency installation, startup hook, or auto-execution mechanism is present; the package contains markdown references, one YAML agent descriptor, and one standard-library Python helper.
Credentials
Network egress is fixed to openapi.dataideaglobal.com, credential use is for HMAC request signing, and local file access is limited to explicit raw request-body input.
Persistence & Privilege
No persistence, background worker, privilege escalation, destructive file operation, broad local indexing, or unrelated credential/session access was found.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install my-skill-test
  3. After installation, invoke the skill by name or use /my-skill-test
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
zingapi-creative-list skill v1.0.0 - Initial release for querying ZingAPI "创意列表" (creative list) endpoint. - Supports parameter conversion from natural language requests to stable API/script calls. - Includes detailed mapping of user inputs to API fields, country code conversion, sorting and filter logic. - Handles default values, credential checks, and error traceability. - Returns grouped summary results with clear business indicator mapping and decoding. - Provides sample usage and parameter reference documentation for seamless adoption.
Metadata
Slug my-skill-test
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is skill-test-demo?

用于调用 ZingAPI 的“创意列表查询”接口。适用于按行业、国家/地区、渠道、关键词、素材类型、AI 标签、时间范围、排序方式等条件检索广告创意、广告素材、创意投放记录、广告主创意和近期新增创意。 It is an AI Agent Skill for Claude Code / OpenClaw, with 39 downloads so far.

How do I install skill-test-demo?

Run "/install my-skill-test" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is skill-test-demo free?

Yes, skill-test-demo is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does skill-test-demo support?

skill-test-demo is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created skill-test-demo?

It is built and maintained by PriestTmob (@priesttomb); the current version is v1.0.0.

💬 Comments