Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

TencentCloud QuestionMark OCR

v1.0.2

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、...

0· 409·2 current·2 all-time
bytencent-ocr@zt1314p-design

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for zt1314p-design/tencentcloud-ocr-questionmarkagent.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "TencentCloud QuestionMark OCR" (zt1314p-design/tencentcloud-ocr-questionmarkagent) from ClawHub.
Skill page: https://clawhub.ai/zt1314p-design/tencentcloud-ocr-questionmarkagent
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Canonical install target

openclaw skills install zt1314p-design/tencentcloud-ocr-questionmarkagent

ClawHub CLI

Package manager switcher

npx clawhub@latest install tencentcloud-ocr-questionmarkagent
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
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.
What to consider before installing
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.

Like a lobster shell, security has layers — review code before you run it.

latestvk97c42errz1rex9wn0k33tw0xh82m7e8
409downloads
0stars
3versions
Updated 19h ago
v1.0.2
MIT-0

腾讯云试题批改Agent (SubmitQuestionMarkAgentJob / DescribeQuestionMarkAgentJob)

用途

调用腾讯云OCR试题批改Agent接口,面向K12教育场景的试题批改产品,支持整卷/单题端到端处理(试卷切题+题目批改+手写坐标回显),聚焦试题批改(含手写答案)和试题解析(不含手写答案)。精准输出题目、正误判定、答案对比、错误分析及知识点等结构化评估结果。低年级算式批改效果优于线上数学作业批改。

核心能力:

  • 异步双接口模式:Submit 提交任务获取 JobId → Describe 轮询查询结果(DONE/FAIL 时完成)
  • 整卷批改:自动切题后逐题批改,返回每道题目的批改信息
  • 单题批改:跳过切题环节,直接对单题进行批改,支持传入参考答案
  • 深度思考:可选开启深度思考模式,进行更深层次推理分析(速度更慢)
  • 知识点输出:可配置输出题目关联的知识点信息
  • 正确答案输出:可配置输出题目的正确答案
  • 手写答案坐标:可配置输出手写答案在原图中的坐标位置
  • 多格式输入:支持 PNG、JPG、JPEG、BMP、GIF、WEBP、HEIC、TIFF、HEIF 及 PDF 格式

官方文档:

使用时机

当用户提出以下需求时触发此技能:

  • 需要对试卷图片/PDF进行自动批改
  • 需要识别手写答案并判断正误
  • 需要对K12试题进行批改分析
  • 需要获取试题的知识点信息
  • 需要获取试题的正确答案
  • 需要对单道题目进行批改(含参考答案)
  • 涉及试卷切题、试题解析的任何场景

环境要求

  • Python 3.6+
  • 依赖:tencentcloud-sdk-python(通过 pip install tencentcloud-sdk-python 安装)
  • 环境变量:
    • TENCENTCLOUD_SECRET_ID:腾讯云API密钥ID
    • TENCENTCLOUD_SECRET_KEY:腾讯云API密钥Key

使用方式

运行 scripts/main.py 脚本完成试题批改。该接口为异步接口,脚本会自动提交任务并轮询查询结果。

请求参数

提交任务 (SubmitQuestionMarkAgentJob) 参数

参数类型必填说明
ImageBase64str否(二选一)图片/PDF的Base64值,编码后不超过10M,分辨率建议600*800以上,支持PNG/JPG/JPEG/BMP/PDF格式
ImageUrlstr否(二选一)图片/PDF的URL地址,下载时间不超过3秒。都提供时只使用ImageUrl。建议存储于腾讯云
PdfPageNumberintPDF页码,仅支持单页识别,默认1
BoolSingleQuestionbool是否单题批改(跳过切题),默认false
EnableDeepThinkbool是否开启深度思考(更深层推理,速度更慢),默认false
QuestionConfigMapstr题目信息输出配置(JSON字符串),可选key:KnowledgePoints(输出知识点)/TrueAnswer(输出正确答案)/ReturnAnswerPosition(输出手写答案坐标)
ReferenceAnswerstr单题批改时的参考答案,仅单题时有效,多题时不生效

查询任务 (DescribeQuestionMarkAgentJob) 参数

参数类型必填说明
JobIdstr任务唯一ID,由Submit接口返回,长度不超过32字符
UserAgentstr请求来源标识(可选),用于追踪调用来源,统一固定为Skills

输出格式

识别成功后返回 JSON 格式结果:

格式化输出模式(默认)

{
  "任务ID": "1410885500986064896",
  "任务状态": "DONE",
  "切题数量": "2",
  "旋转角度": 0.0,
  "批改结果": [
    {
      "题号": 1,
      "题干": "1. 小琪在做作业时发现有一道题的一部分被墨水遮住了...",
      "子题": [
        {
          "题号": "1-1",
          "题干": "(1)小琪猜测,墨水遮住的内容是"2a",请你根据小琪的猜测完成计算;",
          "答案列表": [
            {
              "手写答案": "\\frac{2}{a-1}...",
              "是否正确": false,
              "答案分析": "首先将除法转化为乘法,对分子分母因式分解后约分...",
              "正确答案": "...",
              "知识点": ["分式运算", "因式分解"],
              "答案坐标": [113, 648, 558, 648, 558, 698, 113, 698]
            }
          ]
        }
      ]
    }
  ],
  "RequestId": "xxx"
}

原始输出模式(--raw)

{
  "JobId": "1410885500986064896",
  "JobStatus": "DONE",
  "ErrorCode": "",
  "ErrorMessage": "",
  "Angle": 0.0,
  "MarkInfos": [
    {
      "MarkItemTitle": "1. 小琪在做作业时...",
      "AnswerInfos": [],
      "MarkInfos": [
        {
          "MarkItemTitle": "(1)小琪猜测...",
          "AnswerInfos": [
            {
              "HandwriteInfo": "\\frac{2}{a-1}...",
              "IsCorrect": false,
              "AnswerAnalysis": "首先将除法转化为乘法...",
              "RightAnswer": "",
              "KnowledgePoints": [],
              "HandwriteInfoPositions": [113, 648, 558, 648, 558, 698, 113, 698]
            }
          ],
          "MarkInfos": []
        }
      ]
    }
  ],
  "RequestId": "xxx"
}

响应数据结构说明

SubmitQuestionMarkAgentJob 响应

参数类型说明
JobIdstr任务唯一ID,用于查询结果
QuestionInfolist of QuestionInfo切题题目边框坐标列表(BoolSingleQuestion=true时为空)
QuestionCountstr切题数量,作为计费题目数总量
RequestIdstr唯一请求ID

DescribeQuestionMarkAgentJob 响应

参数类型说明
JobStatusstr任务状态:WAIT(等待中)/RUN(执行中)/FAIL(失败)/DONE(成功)
ErrorCodestr任务执行错误码(非FAIL时为空)
ErrorMessagestr任务执行错误信息(非FAIL时为空)
Anglefloat图片旋转角度(角度制),水平方向为0,顺时针为正
MarkInfoslist of MarkInfo试题批改信息
RequestIdstr唯一请求ID

MarkInfo 结构(嵌套递归):

字段类型说明
MarkItemTitlestr题目的题干信息
AnswerInfoslist of AnswerInfo批改答案列表(按从左到右、从上到下排列)
MarkInfoslist of MarkInfo嵌套子题信息(无子题则为空)

AnswerInfo 结构:

字段类型说明
HandwriteInfostr手写答案内容(如选择题的手写选项、填空题的手写内容)
IsCorrectbool答案是否正确
AnswerAnalysisstr答案分析结果
HandwriteInfoPositionslist of int答案区域4角点坐标(长度8数组:左上/右上/右下/左下),需配置ReturnAnswerPosition。可能返回null
RightAnswerstr正确答案内容,需配置 QuestionConfigMap 的 TrueAnswer 为 true
KnowledgePointslist of str知识点内容,需配置 QuestionConfigMap 的 KnowledgePoints 为 true。可能返回null

QuestionInfo 结构:

字段类型说明
Anglefloat旋转角度
Heightint预处理后图片高度
Widthint预处理后图片宽度
ResultListlist of ResultList文档元素(可能返回null)
OrgHeightint输入图片高度
OrgWidthint输入图片宽度
ImageBase64str预处理后的图片base64编码

错误码说明

错误码含义
FailedOperation.DownLoadError文件下载失败
FailedOperation.ImageDecodeFailed图片解码失败
FailedOperation.ImageSizeTooLarge图片尺寸过大,请确保编码后不超过10M
FailedOperation.OcrFailedOCR识别失败
FailedOperation.PDFParseFailedPDF解析失败
FailedOperation.UnKnowError未知错误
FailedOperation.UnKnowFileTypeError未知的文件类型
FailedOperation.UnOpenError服务未开通,请先在腾讯云控制台开通试题批改Agent服务
InvalidParameterValue.InvalidParameterValueLimit参数值错误
LimitExceeded.TooLargeFileError文件内容太大
ResourceUnavailable.InArrears账号已欠费
ResourceUnavailable.ResourcePackageRunOut账号资源包耗尽
ResourcesSoldOut.ChargeStatusException计费状态异常

重要业务逻辑

  1. 异步双接口模式:Submit 提交任务 → Describe 轮询,JobStatus 为 DONE/FAIL 时完成
  2. Submit 计费,Describe 不计费
  3. ImageBase64 和 ImageUrl 必须提供其一,都提供时只使用 ImageUrl
  4. 支持格式:PNG/JPG/JPEG/BMP/GIF/WEBP/HEIC/TIFF/HEIF 及 PDF
  5. PdfPageNumber 必须为正整数,仅支持单页识别
  6. BoolSingleQuestion=true 时跳过切题,直接单题批改
  7. QuestionConfigMap 示例:{"KnowledgePoints":true,"TrueAnswer":true,"ReturnAnswerPosition":false}
  8. ReferenceAnswer 仅单题时有效,多题时不生效
  9. JobId 长度不超过32字符
  10. 默认接口请求并发限制:10题/分钟
  11. 建议图片存储于腾讯云COS以获得更高的下载速度和稳定性
  12. MarkInfo 为递归嵌套结构,可能存在多层子题

调用示例

# 通过URL进行整卷批改(输出知识点和正确答案)
python scripts/main.py --image-url "https://example.com/exam_paper.jpg" \
  --question-config '{"KnowledgePoints":true,"TrueAnswer":true}'

# 通过URL进行单题批改(带参考答案)
python scripts/main.py --image-url "https://example.com/single_question.jpg" \
  --single-question --reference-answer "x=2"

# 开启深度思考模式
python scripts/main.py --image-url "https://example.com/exam.jpg" --enable-deep-think

# 通过文件路径进行批改(自动Base64编码)
python scripts/main.py --image-base64 ./exam_paper.png

# 批改PDF试卷(指定页码)
python scripts/main.py --image-base64 ./exam.pdf --pdf-page-number 2

# 输出原始JSON响应
python scripts/main.py --image-url "https://example.com/exam.jpg" --raw

# 指定地域和自定义轮询参数
python scripts/main.py --image-url "https://example.com/exam.jpg" \
  --region ap-beijing --poll-interval 3 --poll-timeout 300```

Comments

Loading comments...