← 返回 Skills 市场
TencentCloud QuestionMark OCR
作者
tencent-ocr
· GitHub ↗
· v1.0.2
· MIT-0
432
总下载
0
收藏
2
当前安装
3
版本数
在 OpenClaw 中安装
/install tencentcloud-ocr-questionmarkagent
功能描述
腾讯云试题批改Agent(SubmitQuestionMarkAgentJob/DescribeQuestionMarkAgentJob)接口调用技能。当用户需要对试卷图片或试题图片中的K12试卷或试题进行自动批改、手写答案识别、知识点分析时,应使用此技能。支持整卷图片批改和单题图片批改,提供题目切题、正误判定、...
安全使用建议
This skill implements Tencent Cloud's async OCR/marking API and will send provided images to Tencent Cloud for processing. Before installing: (1) Confirm you are comfortable sending exam/answer images to Tencent Cloud (privacy/billing implications). (2) Provide dedicated Tencent Cloud API keys with minimal permissions and monitor usage/billing; do not reuse high-privilege or personal root keys. (3) Install the Python dependency (tencentcloud-sdk-python) in a controlled environment before running the script. (4) Note the registry metadata does not declare the required environment variables — ask the publisher to correct that omission or verify the skill's manifest. (5) If you want to audit behavior, review scripts/main.py fully (it only encodes/reads images, calls Tencent Cloud SDK, and formats results). If you are not comfortable exposing images or API keys, do not install or run this skill.
功能分析
Type: OpenClaw Skill
Name: tencentcloud-ocr-questionmarkagent
Version: 1.0.2
The skill bundle is a legitimate implementation for interacting with the Tencent Cloud OCR Question Mark Agent API. The Python script (scripts/main.py) uses the official tencentcloud-sdk-python to submit and poll for exam grading tasks, correctly handling environment variables for credentials and implementing standard file-to-base64 conversion logic without any signs of malicious intent, data exfiltration, or unauthorized execution.
能力评估
Purpose & Capability
The skill's name, description, SKILL.md and scripts/main.py consistently implement Tencent Cloud's SubmitQuestionMarkAgentJob/DescribeQuestionMarkAgentJob flow (submitting an async OCR/batch-marking job and polling for results). Requiring Tencent Cloud API keys (TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY) and the tencentcloud-sdk-python dependency is coherent with the stated purpose. However, the registry metadata at the top of the skill report lists no required environment variables or primary credential — that is inconsistent with both the SKILL.md and the code.
Instruction Scope
SKILL.md and the script limit actions to: reading a provided image file (or base64 string), encoding it if needed, calling Tencent Cloud APIs via the official SDK, and polling for job results. There are no instructions to read unrelated system files or to send data to third-party endpoints beyond Tencent Cloud. The script will read local files when you pass a filepath for the image. The notable issue is that the runtime instructions explicitly require Tencent Cloud credentials, but the registry metadata omitted that requirement.
Install Mechanism
No install spec is provided (instruction-only with an included script). SKILL.md lists a dependency on 'tencentcloud-sdk-python' but the skill does not provide an automated install step. This is low technical risk but operationally important: the runtime will fail unless the SDK is installed. No high-risk downloads or obscure URLs are present.
Credentials
The code requires TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY to call Tencent Cloud APIs — these are appropriate and proportional for the service. The concern is that the published skill metadata omits these required credentials, which is misleading and could result in unexpected prompts or manual configuration. Because these credentials can be used for billing and API access, the skill should have explicitly declared them in the registry metadata and documented least-privilege recommendations.
Persistence & Privilege
The skill does not request permanent/always-on inclusion, does not modify other skills or system-wide agent settings, and does not request elevated privileges. It only performs on-demand network calls to Tencent Cloud when invoked.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install tencentcloud-ocr-questionmarkagent - 安装完成后，直接呼叫该 Skill 的名称或使用
/tencentcloud-ocr-questionmarkagent触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.2
- 新增 UserAgent 的说明，可用于追踪请求来源，固定为 Skills
- 其他功能及接口未变，保持与上一版本一致
v1.0.1
更新显示名称
v1.0.0
- Initial release of the Tencent Cloud OCR QuestionMarkAgent skill.
- Supports automatic grading, handwritten answer recognition, and knowledge point analysis for K12 exam or question images (whole sheet and single-question modes).
- Implements asynchronous job submission and result polling (SubmitQuestionMarkAgentJob / DescribeQuestionMarkAgentJob API).
- Provides configurable output for knowledge points, correct answers, answer coordinates, and deep thinking mode.
- Accepts various image and PDF formats via URL or Base64 input.
- Includes sample commands and detailed error code documentation.
元数据
常见问题
TencentCloud QuestionMark OCR 是什么？
腾讯云试题批改Agent(SubmitQuestionMarkAgentJob/DescribeQuestionMarkAgentJob)接口调用技能。当用户需要对试卷图片或试题图片中的K12试卷或试题进行自动批改、手写答案识别、知识点分析时,应使用此技能。支持整卷图片批改和单题图片批改,提供题目切题、正误判定、... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 432 次。
如何安装 TencentCloud QuestionMark OCR？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install tencentcloud-ocr-questionmarkagent」即可一键安装，无需额外配置。
TencentCloud QuestionMark OCR 是免费的吗？
是的，TencentCloud QuestionMark OCR 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
TencentCloud QuestionMark OCR 支持哪些平台？
TencentCloud QuestionMark OCR 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 TencentCloud QuestionMark OCR？
由 tencent-ocr（@zt1314p-design）开发并维护，当前版本 v1.0.2。
推荐 Skills