Zyt ai creation

Use Chanjing AI creation APIs to submit image or video generation tasks across multiple models, inspect task status, poll async results, and explicitly downl...

Security Scan

The skill's code, instructions, and resource access line up with its stated purpose (submitting/polling Chanjing AI creation tasks); it reads/writes a local credentials file and talks to chanjing domains, which is expected for this integration.

Purpose & Capability

Name/description match the included scripts and docs: submit_task, poll_task, get_task, list_tasks and optional download_result. The scripts call Chanjing API endpoints and use local app_id/secret_key — coherent with an API client for Chanjing AI creation.

Instruction Scope

SKILL.md instructs only to submit tasks, poll status, list/get tasks, and download results only when explicitly requested. The included scripts read/write ~/.chanjing/credentials.json (or $CHANJING_CONFIG_DIR/credentials.json) and may open the official login page in the browser when credentials are missing/invalid — this behavior is documented and consistent with the skill purpose.

Install Mechanism

No install spec; this is an instruction-only skill with two small Python helper modules. Nothing is downloaded from arbitrary URLs and no installer runs, which is low-risk and proportionate.

Credentials

The skill does not require external credentials from the registry metadata; it expects the user to provide Chanjing app_id/secret_key in ~/.chanjing/credentials.json (or via CHANJING_CONFIG_DIR). It optionally respects CHANJING_API_BASE if set. This is appropriate, but note it stores access_token and expire_in back into the same credentials file in plaintext — users should be aware of local storage of secrets.

Persistence & Privilege

always is false and the skill is user-invocable; it only writes/reads its own config file (~/.chanjing/credentials.json) and does not modify other skills or system-wide agent settings. It may open the default browser for login flows as documented.

Assessment

This skill appears to be a straightforward Chanjing AI creation client. Before installing, verify you trust the chanjing endpoints (https://open-api.chanjing.cc and https://www.chanjing.cc), and be prepared to store your app_id and secret_key in ~/.chanjing/credentials.json (or set CHANJING_CONFIG_DIR). The scripts will cache access_token and expire_in in that file in plaintext — if that is a concern, keep the directory protected or avoid storing credentials there. Also avoid setting CHANJING_API_BASE to an untrusted URL (the code honors that env var). If you want extra caution, inspect the two Python files (scripts/_auth.py and scripts/_task_api.py) before use and only allow downloads when you explicitly request saving results locally.

Chanjing AI Creation

When to Use This Skill

当用户要做这些事时使用本 Skill：

• 用蝉镜 AI 创作模型生成图片
• 用蝉镜 AI 创作模型生成视频
• 查询 AI 创作任务列表或单个任务详情
• 轮询 AI 创作异步结果
• 在用户明确要求时下载图片或视频到本地

如果需求更接近“文生数字人”，优先使用 chanjing-text-to-digital-person。
如果需求更接近“已有数字人视频合成”，优先使用 chanjing-video-compose。

Preconditions

本 Skill 自己包含本地配置和鉴权流程，不依赖其他 skill 的运行时脚本。

本 Skill 使用：

• 配置文件：~/.chanjing/credentials.json
• 若设置环境变量 CHANJING_CONFIG_DIR：使用 $CHANJING_CONFIG_DIR/credentials.json
• API 基础地址：https://open-api.chanjing.cc（可用 CHANJING_API_BASE 覆盖）

当本地缺少 AK/SK 或 AK/SK 无效时，脚本可能在默认浏览器打开蝉镜官网登录页：
https://www.chanjing.cc/openapi/login

Standard Workflow

AI 创作的主接口是统一提交器：

1. 调用 submit_task 提交图片或视频生成任务，得到 unique_id
2. 调用 poll_task 轮询直到成功，得到 output_url
3. 如需回看任务参数或错误原因，调用 get_task
4. 如需看历史记录，调用 list_tasks
5. 只有在用户明确要求保存到本地时，才调用 download_result

这个 skill 默认做成“通用任务提交器”：

• 对常见图片/视频模型，优先使用脚本提供的通用参数
• 对特殊模型参数，使用 --body-file 或 --body-json 透传完整请求体

Covered APIs

本 Skill 当前覆盖：

• POST /open/v1/ai_creation/task/submit
• POST /open/v1/ai_creation/task/page
• GET /open/v1/ai_creation/task

Scripts

脚本目录：

• scripts/

脚本	说明
chanjing-config	写入/查看本地 app_id 与 secret_key，并清理旧 token 缓存
chanjing-get-token	从本地凭证获取有效 access_token（必要时自动刷新）
_auth.py	读取本地凭证、获取或刷新 access_token
submit_task	提交 AI 创作任务，输出 unique_id
get_task	获取单个任务详情
list_tasks	列出图片或视频任务
poll_task	轮询任务直到完成，默认输出第一个结果地址
download_result	下载图片或视频到 outputs/ai-creation/

Usage Examples

示例 1：Seedream 3.0 文生图

TASK_ID=$(python3 scripts/submit_task \
  --creation-type 3 \
  --model-code "doubao-seedream-3.0-t2i" \
  --prompt "赛博朋克城市夜景，霓虹灯，雨夜，电影镜头" \
  --aspect-ratio "16:9" \
  --clarity 2048 \
  --number-of-images 1)

python3 scripts/poll_task --unique-id "$TASK_ID"

示例 2：腾讯 Kling v2.1 Master 图生视频

TASK_ID=$(python3 scripts/submit_task \
  --creation-type 4 \
  --model-code "tx_kling-v2-1-master" \
  --ref-img-url "https://res.chanjing.cc/chanjing/res/aigc_creation/photo/start.jpg" \
  --ref-img-url "https://res.chanjing.cc/chanjing/res/aigc_creation/photo/end.jpg" \
  --prompt "角色从静止到转身，镜头平滑移动，叙事感强" \
  --aspect-ratio "9:16" \
  --clarity 1080 \
  --quality-mode pro \
  --video-duration 5)

python3 scripts/poll_task --unique-id "$TASK_ID"

示例 3：直接透传完整 JSON

python3 scripts/submit_task \
  --body-file ./payload.json

Download Rule

下载是显式动作，不是默认动作：

• poll_task 成功后应先返回远端 output_url
• 不要自动下载结果文件
• 只有当用户明确表达“下载到本地”“保存到 outputs”“帮我落盘”时，才执行 download_result

Output Convention

默认本地输出目录：

• outputs/ai-creation/

Additional Resources

更多接口细节见：

• skills/chanjing-ai-creation/reference.md
• skills/chanjing-ai-creation/examples.md