@seacloudai/seacloud-cli 0.0.9 → 0.0.11

package/README.md

@@ -28,6 +28,7 @@
 - **Authentication**: Sign in with the browser-based device flow and store credentials locally.
 - **Model discovery**: List available models and inspect full parameter specs in human-readable or JSON form.
 - **Task execution**: Submit multimodal generation tasks from the CLI with parameter validation and structured output options.
+- **Image generation via proxy**: Call sync image-generation models through `folkos-proxy`, with optional asset URL output.
 - **Task tracking**: Poll task status and print result URLs or full JSON responses.
 - **SkillHub integration**: Search, install, and configure agent skills from SeaCloud SkillHub.
 - **Agent-friendly UX**: Supports `--dry-run`, JSON output, stable command shapes, and copy-pasteable examples.
@@ -92,16 +93,33 @@ seacloud auth status
 ```bash
 seacloud models list
 seacloud models list --type video
-seacloud models spec kirin_v2_6_i2v
-seacloud models spec kirin_v2_6_i2v --output json
+seacloud models spec kling_v2_6_i2v
+seacloud models spec seedance_2_0 --output json
 ```
 
 ### Run a task
 
 ```bash
-seacloud run kirin_v2_6_i2v --param image=https://example.com/cat.jpg
-seacloud run kirin_v2_6_i2v --param prompt="a cat running" --param duration=5
-seacloud run kirin_v2_6_i2v --param mode=pro --output url
+seacloud run kling_v2_6_i2v --param image=https://example.com/cat.jpg
+seacloud run seedance_2_0 --param prompt="a cat running" --param duration=5
+seacloud run kling_v2_6_i2v --param mode=pro --output url
+seacloud run gpt-image-2 --param prompt="a blue cat" --output url
+```
+
+SeaCloud CLI accepts user-facing model IDs such as `kling_*`, `seedance_*`, and `seedream_*`.
+They are resolved to the current backend IDs automatically before submission.
+
+### Generate an image through the proxy
+
+```bash
+SEACLOUD_FOLKOS_PROXY_URL=http://127.0.0.1:8090 seacloud images generate \
+  --model gpt-image-2 \
+  --prompt "A flat vector poster of a blue cat wearing black sunglasses" \
+  --output json
+
+SEACLOUD_FOLKOS_PROXY_URL=http://127.0.0.1:8090 seacloud images generate \
+  --prompt "A flat vector poster of a blue cat wearing black sunglasses" \
+  --output url
 ```
 
 ### Check task status
@@ -148,6 +166,7 @@ seacloud models spec <model_id> --output json
 seacloud run <model_id> --param key=value
 seacloud run <model_id> --param prompt="hello" --param duration=5
 seacloud run <model_id> --output json
+seacloud run gpt-image-2 --param prompt="a blue cat" --output url
 ```
 
 Nested fields use dot notation:
@@ -173,6 +192,14 @@ seacloud skills add <slug>
 seacloud skills config --show
 ```
 
+### `seacloud images`
+
+```bash
+seacloud images generate --prompt="a blue cat"
+seacloud images generate --prompt="a blue cat" --output json
+seacloud images generate --prompt="a blue cat" --output url
+```
+
 ### `seacloud version`
 
 ```bash
@@ -183,12 +210,13 @@ seacloud version
 
 - Use `--output json` where supported for machine-readable responses.
 - Use `--output url` on task commands to print only result URLs.
+- Set `SEACLOUD_FOLKOS_PROXY_URL` to the root of your `folkos-proxy` service when using `seacloud images generate` or sync image models through `seacloud run`.
 - Use global `--dry-run` to inspect execution without sending requests.
 
 Example:
 
 ```bash
-seacloud --dry-run run kirin_v2_6_i2v --param prompt=test
+seacloud --dry-run run seedance_2_0 --param prompt=test
 ```
 
 ## Release

package/README.zh.md

@@ -28,6 +28,7 @@
 - **认证登录**：支持浏览器设备码登录，并将凭证安全保存在本地。
 - **模型发现**：列出可用模型，并以可读文本或 JSON 查看完整参数规格。
 - **任务执行**：通过 CLI 提交多模态生成任务，支持参数校验和结构化输出。
+- **代理生图**：通过 `folkos-proxy` 调用同步生图模型，并可选择输出资产 URL。
 - **任务追踪**：轮询任务状态，输出结果 URL 或完整 JSON。
 - **SkillHub 集成**：搜索、安装和配置 SeaCloud SkillHub 技能。
 - **Agent 友好**：支持 `--dry-run`、JSON 输出、稳定命令结构和可直接复制的示例。
@@ -92,16 +93,33 @@ seacloud auth status
 ```bash
 seacloud models list
 seacloud models list --type video
-seacloud models spec kirin_v2_6_i2v
-seacloud models spec kirin_v2_6_i2v --output json
+seacloud models spec kling_v2_6_i2v
+seacloud models spec seedance_2_0 --output json
 ```
 
 ### 执行任务
 
 ```bash
-seacloud run kirin_v2_6_i2v --param image=https://example.com/cat.jpg
-seacloud run kirin_v2_6_i2v --param prompt="a cat running" --param duration=5
-seacloud run kirin_v2_6_i2v --param mode=pro --output url
+seacloud run kling_v2_6_i2v --param image=https://example.com/cat.jpg
+seacloud run seedance_2_0 --param prompt="a cat running" --param duration=5
+seacloud run kling_v2_6_i2v --param mode=pro --output url
+seacloud run gpt-image-2 --param prompt="一只蓝色猫" --output url
+```
+
+SeaCloud CLI 现在支持直接使用 `kling_*`、`seedance_*`、`seedream_*` 这类对外模型 ID，
+提交前会自动映射到当前后端使用的原始模型 ID。
+
+### 通过代理生图
+
+```bash
+SEACLOUD_FOLKOS_PROXY_URL=http://127.0.0.1:8090 seacloud images generate \
+  --model gpt-image-2 \
+  --prompt "A flat vector poster of a blue cat wearing black sunglasses" \
+  --output json
+
+SEACLOUD_FOLKOS_PROXY_URL=http://127.0.0.1:8090 seacloud images generate \
+  --prompt "A flat vector poster of a blue cat wearing black sunglasses" \
+  --output url
 ```
 
 ### 查询任务状态
@@ -148,6 +166,7 @@ seacloud models spec <model_id> --output json
 seacloud run <model_id> --param key=value
 seacloud run <model_id> --param prompt="hello" --param duration=5
 seacloud run <model_id> --output json
+seacloud run gpt-image-2 --param prompt="一只蓝色猫" --output url
 ```
 
 嵌套字段支持 dot notation：
@@ -173,6 +192,14 @@ seacloud skills add <slug>
 seacloud skills config --show
 ```
 
+### `seacloud images`
+
+```bash
+seacloud images generate --prompt="一只蓝色猫"
+seacloud images generate --prompt="一只蓝色猫" --output json
+seacloud images generate --prompt="一只蓝色猫" --output url
+```
+
 ### `seacloud version`
 
 ```bash
@@ -183,12 +210,13 @@ seacloud version
 
 - 在支持的命令上使用 `--output json` 获取机器可读输出。
 - 在任务命令上使用 `--output url` 只打印结果 URL。
+- 使用 `seacloud images generate` 或通过 `seacloud run` 调用同步生图模型时，请把 `SEACLOUD_FOLKOS_PROXY_URL` 设置为 `folkos-proxy` 服务根地址。
 - 使用全局 `--dry-run` 在不发请求的前提下检查执行内容。
 
 示例：
 
 ```bash
-seacloud --dry-run run kirin_v2_6_i2v --param prompt=test
+seacloud --dry-run run seedance_2_0 --param prompt=test
 ```
 
 ## 发布说明

package/npm-bundles/SHA256SUMS

@@ -1,5 +1,5 @@
-198f18eb05b08b1474e8d4ee3a27229ad9d831dce7d8576cf8cc6ebda295500c  seacloud_0.0.9_darwin_amd64.tar.gz
-1d26dc477513086afc13b49542325aa473efab91905d138a6a0a7398b8533198  seacloud_0.0.9_darwin_arm64.tar.gz
-d883e69a63c8a0ac6e0c7de6a178b37afc727ac46b0f3ae29e367a06a2bcd340  seacloud_0.0.9_linux_amd64.tar.gz
-9ff1c5f9fe0547ee95d930dde8ca3c07eadc7f092c4013b1433536640a2f5042  seacloud_0.0.9_linux_arm64.tar.gz
-bad39e30b8e33183c47d1fbdebf574a47784a7cbb5b54a510122c257ebefb1eb  seacloud_0.0.9_windows_amd64.zip
+3a5c642635a7acd5e597b20080c37dad88883c0ca31c2da11dafa069077329f7  seacloud_0.0.11_darwin_amd64.tar.gz
+0d1990645e3af0465dada237256398677b64cfe894cd58246caba599fdec98c6  seacloud_0.0.11_darwin_arm64.tar.gz
+5e9c64e24b43ea0fad6265a48aeb744dab8bae8b398d2bc59ea4b257d0879d5d  seacloud_0.0.11_linux_amd64.tar.gz
+c9236b1b4e0bf07c4c6258d4278c844f457d35b2ebad8d4d755169eddc345d06  seacloud_0.0.11_linux_arm64.tar.gz
+56077615a83f5dcfd7adfde6cf2a5f86d364b6358b54c84860137a15cb37be83  seacloud_0.0.11_windows_amd64.zip

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@seacloudai/seacloud-cli",
-  "version": "0.0.9",
+  "version": "0.0.11",
   "description": "npm installer wrapper for the SeaCloud CLI binary",
   "license": "MIT",
   "publishConfig": {