PaddlePaddle · Bobholamovic · Apr 20, 2026 · Mar 14, 2026 · Mar 17, 2026 · Mar 24, 2026
diff --git a/skills/README.md b/skills/README.md
@@ -9,8 +9,8 @@ This directory contains official PaddleOCR Agent Skills. They integrate with AI
 
 ## Prerequisites
 
-1. Python 3.8 or later must be installed on the device that runs the skill.
-2. These skills depend on PaddleOCR official APIs and require API credentials. Visit the [PaddleOCR website](https://www.paddleocr.com), click **API**, select the model you need, then copy the `API_URL` and `Token`. They correspond to the API URL and access token used for authentication. Supported models per skill:
+1. Python 3.9 or later and [uv](https://docs.astral.sh/uv/) must be installed on the device that runs the skill. Scripts declare dependencies inline ([PEP 723](https://peps.python.org/pep-0723/)); `uv run` resolves them automatically.
+2. These skills depend on PaddleOCR official APIs and require API credentials. Visit the [PaddleOCR website](https://www.paddleocr.com), click **API**, select the model you need, select the language for the text recognition model, then copy the `API_URL` and `Token`. They correspond to the API URL and access token used for authentication. Supported model per skill:
    - `paddleocr-text-recognition`: `PP-OCRv5`
    - `paddleocr-doc-parsing`: `PP-StructureV3`, `PaddleOCR-VL`, `PaddleOCR-VL-1.5`
 
@@ -34,6 +34,7 @@ npx skills add PaddlePaddle/PaddleOCR -g --skill paddleocr-doc-parsing -y
 > ```shell
 > git clone https://github.com/PaddlePaddle/PaddleOCR.git
 > npx skills add ./PaddleOCR/skills/paddleocr-text-recognition
+> npx skills add ./PaddleOCR/skills/paddleocr-doc-parsing
 > ```
 
 #### Option 2: Install via `clawhub` (OpenClaw)
@@ -65,8 +66,8 @@ After installation, configure the required environment variables so the skills c
 
 | Skill | Required | Optional |
 | --- | --- | --- |
-| `paddleocr-text-recognition` | `PADDLEOCR_OCR_API_URL` (API URL), `PADDLEOCR_ACCESS_TOKEN` (access token) | `PADDLEOCR_OCR_TIMEOUT` (API request timeout) |
-| `paddleocr-doc-parsing` | `PADDLEOCR_DOC_PARSING_API_URL` (API URL), `PADDLEOCR_ACCESS_TOKEN` (access token) | `PADDLEOCR_DOC_PARSING_TIMEOUT` (API request timeout) |
+| `paddleocr-text-recognition` | `PADDLEOCR_OCR_API_URL` (full endpoint URL ending with `/ocr`), `PADDLEOCR_ACCESS_TOKEN` (access token) | `PADDLEOCR_OCR_TIMEOUT` (API request timeout) |
+| `paddleocr-doc-parsing` | `PADDLEOCR_DOC_PARSING_API_URL` (full endpoint URL ending with `/layout-parsing`), `PADDLEOCR_ACCESS_TOKEN` (access token) | `PADDLEOCR_DOC_PARSING_TIMEOUT` (API request timeout) |
 
 Below are configuration methods for some AI apps:
 
@@ -145,28 +146,21 @@ This section describes how to run smoke tests locally to verify that the skills
 
 > The examples below cover both skills. Run only the commands for the skill(s) you need.
 
-Make sure your working directory is the directory containing this file.
+Make sure your working directory is the directory containing this file. All scripts declare their dependencies inline ([PEP 723](https://peps.python.org/pep-0723/)), so [uv](https://docs.astral.sh/uv/) resolves them automatically — no separate install step is needed.
 
-1. Install dependencies.
-
-   ```shell
-   python -m pip install -r paddleocr-text-recognition/scripts/requirements.txt
-   python -m pip install -r paddleocr-doc-parsing/scripts/requirements.txt
-   # Optional: required only when using document file optimization
-   python -m pip install -r paddleocr-doc-parsing/scripts/requirements-optimize.txt
-   ```
-
-2. Configure environment variables (see [Configure Environment Variables](#configure-environment-variables) for the list of variables).
+1. Configure environment variables (see [Configure Environment Variables](#configure-environment-variables) for the list of variables).
 
    ```shell
    export PADDLEOCR_OCR_API_URL="<OCR_API_URL>"
    export PADDLEOCR_ACCESS_TOKEN="<ACCESS_TOKEN>"
    export PADDLEOCR_DOC_PARSING_API_URL="<DOC_PARSING_API_URL>"
    ```
 
-3. Run the smoke test scripts.
+2. Run the smoke test scripts.
 
    ```shell
-   python paddleocr-text-recognition/scripts/smoke_test.py
-   python paddleocr-doc-parsing/scripts/smoke_test.py
+   cd paddleocr-text-recognition && uv run scripts/smoke_test.py && cd ..
+   cd paddleocr-doc-parsing && uv run scripts/smoke_test.py && cd ..
    ```
+
+   Use `--skip-api-test` to verify configuration only (no network call). Use `--test-url "https://..."` to override the default sample document/image URL.
diff --git a/skills/README_cn.md b/skills/README_cn.md
@@ -9,8 +9,8 @@
 
 ## 准备工作
 
-1. 请确保执行 skill 的设备安装有 Python 3.8 或以上版本。
-2. Skill 底层依赖于 PaddleOCR 官方 API，因此需要配置相关凭证才能使用。可以在 [PaddleOCR 官网](https://www.paddleocr.com) 点击 **API**，选择需要用到的算法，然后复制 `API_URL` 和 `Token`，它们分别对应服务的 API URL 和用户鉴权使用的 access token。各 skill 支持的算法如下：
+1. 请确保执行 skill 的设备安装有 Python 3.9 或以上版本和 [uv](https://docs.astral.sh/uv/)。所有脚本均以 [PEP 723](https://peps.python.org/pep-0723/) 格式内联声明依赖，`uv run` 会自动解析。
+2. Skill 底层依赖于 PaddleOCR 官方 API，因此需要配置相关凭证才能使用。可以在 [PaddleOCR 官网](https://www.paddleocr.com) 点击 **API**，选择需要用到的模型，选择语言（对于文字识别模型），然后复制 `API_URL` 和 `Token`，它们分别对应服务的 API URL 和用户鉴权使用的 access token。各 skill 支持的模型如下：
    - `paddleocr-text-recognition`：`PP-OCRv5`
    - `paddleocr-doc-parsing`：`PP-StructureV3`、`PaddleOCR-VL`、`PaddleOCR-VL-1.5`
 
@@ -34,6 +34,7 @@ npx skills add PaddlePaddle/PaddleOCR -g --skill paddleocr-doc-parsing -y
 > ```shell
 > git clone https://github.com/PaddlePaddle/PaddleOCR.git
 > npx skills add ./PaddleOCR/skills/paddleocr-text-recognition
+> npx skills add ./PaddleOCR/skills/paddleocr-doc-parsing
 > ```
 
 #### 方式二：通过 `clawhub` 安装（OpenClaw）
@@ -65,8 +66,8 @@ git clone https://github.com/PaddlePaddle/PaddleOCR.git
 
 | Skill | 必填 | 可选 |
 | --- | --- | --- |
-| `paddleocr-text-recognition` | `PADDLEOCR_OCR_API_URL`（API URL）、`PADDLEOCR_ACCESS_TOKEN`（access token） | `PADDLEOCR_OCR_TIMEOUT`（API 请求超时时间） |
-| `paddleocr-doc-parsing` | `PADDLEOCR_DOC_PARSING_API_URL`（API URL）、`PADDLEOCR_ACCESS_TOKEN`（access token） | `PADDLEOCR_DOC_PARSING_TIMEOUT`（API 请求超时时间） |
+| `paddleocr-text-recognition` | `PADDLEOCR_OCR_API_URL`（完整端点 URL，须以 `/ocr` 结尾）、`PADDLEOCR_ACCESS_TOKEN`（access token） | `PADDLEOCR_OCR_TIMEOUT`（API 请求超时时间） |
+| `paddleocr-doc-parsing` | `PADDLEOCR_DOC_PARSING_API_URL`（完整端点 URL，须以 `/layout-parsing` 结尾）、`PADDLEOCR_ACCESS_TOKEN`（access token） | `PADDLEOCR_DOC_PARSING_TIMEOUT`（API 请求超时时间） |
 
 以下是部分 AI 应用的配置方式：
 
@@ -145,28 +146,21 @@ git clone https://github.com/PaddlePaddle/PaddleOCR.git
 
 > 以下示例覆盖多个 skill。如果只需使用某一个 skill，只需执行该 skill 对应的命令。
 
-执行前，请确保工作目录位于本文档所在的目录下。
+执行前，请确保工作目录位于本文档所在的目录下。所有脚本均以 [PEP 723](https://peps.python.org/pep-0723/) 格式内联声明依赖，[uv](https://docs.astral.sh/uv/) 会自动解析——无需单独安装依赖。
 
-1. 安装依赖库。
-
-   ```shell
-   python -m pip install -r paddleocr-text-recognition/scripts/requirements.txt
-   python -m pip install -r paddleocr-doc-parsing/scripts/requirements.txt
-   # 可选依赖，仅在优化文档文件大小时需要
-   python -m pip install -r paddleocr-doc-parsing/scripts/requirements-optimize.txt
-   ```
-
-2. 配置环境变量（需要配置的变量参见[配置环境变量](#配置环境变量)一节）。
+1. 配置环境变量（需要配置的变量参见[配置环境变量](#配置环境变量)一节）。
 
    ```shell
    export PADDLEOCR_OCR_API_URL="<OCR_API_URL>"
    export PADDLEOCR_ACCESS_TOKEN="<ACCESS_TOKEN>"
    export PADDLEOCR_DOC_PARSING_API_URL="<DOC_PARSING_API_URL>"
    ```
 
-3. 运行冒烟测试脚本。
+2. 运行冒烟测试脚本。
 
    ```shell
-   python paddleocr-text-recognition/scripts/smoke_test.py
-   python paddleocr-doc-parsing/scripts/smoke_test.py
+   cd paddleocr-text-recognition && uv run scripts/smoke_test.py && cd ..
+   cd paddleocr-doc-parsing && uv run scripts/smoke_test.py && cd ..
    ```
+
+   使用 `--skip-api-test` 可只做配置检查（不发网络请求）。使用 `--test-url "https://..."` 可指定自定义测试用文档/图片 URL。