← 返回 Skills 市场
Child Learning Behavior Analysis Tool | 孩子学习行为分析工具
作者
smyx-sunjinhui
· GitHub ↗
· v1.0.1
· MIT-0
104
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install smyx-study-analysis
功能描述
Conducts video analysis of learning behavior for children/students, identifies poor learning habits, provides structured analysis reports and family educatio...
安全使用建议
Before installing or enabling this skill consider the following:
- The skill will accept videos and upload them to remote analysis APIs. Default production endpoints are present in the bundled config (lifeemergence/open.lifeemergence.com). If you do not trust that endpoint, do not upload sensitive videos.
- The manifest declared no required env vars or config files, but the skill reads workspace config files and environment variables (OPENCLAW_WORKSPACE, OPENCLAW_SENDER_OPEN_ID, etc.) and will create/use a local SQLite database under the workspace. Expect persistent storage of uploaded videos and reports in the agent workspace/data and attachments folders.
- The package bundles a large common library (smyx_common) and an unrelated face_analysis module; this increases the attack surface and means the skill can access generic utilities (HTTP, DB, filesystem) beyond narrow video-analysis code.
- If you proceed: (1) review/override the skill config to point to an endpoint you control or a vetted service; (2) run the skill in an isolated workspace (not containing other sensitive files) and set OPENCLAW_WORKSPACE to an empty folder; (3) do not enable it to process videos containing private or identifying content unless you accept the upload to the remote server; (4) verify where attachments and DB files will be stored and periodically inspect/clean them.
Because of the mismatches between declared metadata and actual behavior (filesystem writes, env reads, external uploads, extra modules), treat this skill as suspicious and only enable it after you have validated the endpoint and workspace configuration.
功能分析
Type: OpenClaw Skill
Name: smyx-study-analysis
Version: 1.0.1
The skill bundle provides a comprehensive tool for analyzing children's learning behavior via video, utilizing a cloud-based API (lifeemergence.com). It includes a robust shared utility library (smyx_common) that handles authentication, local token caching in a SQLite database, and structured API communication. The SKILL.md file contains strict 'Forced Memory Rules' designed to ensure the AI agent retrieves historical data exclusively from the cloud API rather than potentially stale local memory files, which serves as a functional constraint rather than a malicious injection. No evidence of unauthorized data exfiltration, malicious execution, or deceptive behavior was found; the code is well-structured and aligned with its stated educational purpose.
能力标签
能力评估
Purpose & Capability
The declared purpose (video analysis of children's study behavior) is consistent with code that validates video files and calls remote analysis APIs. However the package includes a large shared 'smyx_common' library (DB/DAO/config utilities) and an unrelated 'face_analysis' skill set (TCM face diagnosis) bundled together. The manifest claims no required config paths or env vars, but SKILL.md and the code expect and read config files and environment variables (e.g. skills/smyx_common/scripts/config.yaml, OPENCLAW_WORKSPACE and OPENCLAW_SENDER_OPEN_ID). That mismatch between claimed requirements and actual dependencies is a design inconsistency.
Instruction Scope
Runtime instructions require obtaining an open-id (by reading skills/smyx_common/scripts/config.yaml under the skill directory or a workspace-level config), saving uploaded attachments into an attachments directory, and running scripts that will send files to a remote API. The SKILL.md explicitly forbids reading local memory files, but code and utilities will read and create YAML config files and a workspace SQLite database; the skill also instructs automatic saving of uploaded videos to disk. The instructions give the agent explicit permission/requirements to access filesystem locations and to call external endpoints — more breadth than the registry metadata declares.
Install Mechanism
There is no install spec (instruction-only flag), but the package contains many Python modules and requirements lists (skills/smyx_common/requirements.txt and face_analysis/requirements.txt) implying heavy dependencies. Because no installer is declared, these dependencies would not be automatically installed by the platform; nevertheless the large dependency set and included native-style code increase the skill's runtime footprint compared to a pure instruction-only skill.
Credentials
Skill metadata declares no required environment variables or config paths, but the code reads environment variables (OPENCLAW_SENDER_OPEN_ID, OPENCLAW_SENDER_USERNAME, FEISHU_OPEN_ID, OPENCLAW_WORKSPACE) and expects YAML config files supplying API base URLs and API keys. The skill will upload videos and POST data to remote endpoints (ApiEnum base URLs from smyx_common config, which default to production domains like open.lifeemergence.com / lifeemergence.com). Requesting or implicitly using these workspace/env values without declaring them is disproportionate and increases risk of unintended data exfiltration.
Persistence & Privilege
The code includes a DAO that creates/uses a SQLite DB under a data directory inside the OpenClaw workspace (derived from OPENCLAW_WORKSPACE or by traversing parent directories). It will also save uploaded attachments to an attachments directory. The skill therefore persists user-submitted videos and analysis results on disk and can create config files if missing. While `always` is false, the skill still has the ability to write persistent files and databases in the agent workspace — a notable privilege not declared in the registry metadata.
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 smyx-study-analysis - 安装完成后，直接呼叫该 Skill 的名称或使用
/smyx-study-analysis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
v1.0.1
- Updated version to 1.0.1.
- Multiple scripts and configuration files modified, including config.py and config.yaml.
- No functional or documentation changes in SKILL.md; content remains unchanged from previous version.
- Internal adjustments likely for bug fixes, compatibility, or configuration improvements.
v1.0.0
- Initial release of the study-analysis skill for evaluating children's/student's learning behaviors via video analysis.
- Identifies poor study habits (e.g., improper posture, distraction, fidgeting) and generates structured reports with improvement suggestions.
- Enforces strict data access rules: all historical report queries must use the cloud API; local memory sources are strictly forbidden.
- Requires open-id acquisition via a controlled workflow before any analysis; execution is paused and prompts the user if open-id is missing.
- Outputs standardized Markdown tables for report listings, including direct report access links.
- Provides parents with actionable family education suggestions grounded in educational psychology principles.
元数据
常见问题
Child Learning Behavior Analysis Tool | 孩子学习行为分析工具 是什么？
Conducts video analysis of learning behavior for children/students, identifies poor learning habits, provides structured analysis reports and family educatio... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 104 次。
如何安装 Child Learning Behavior Analysis Tool | 孩子学习行为分析工具？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install smyx-study-analysis」即可一键安装，无需额外配置。
Child Learning Behavior Analysis Tool | 孩子学习行为分析工具 是免费的吗？
是的，Child Learning Behavior Analysis Tool | 孩子学习行为分析工具 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Child Learning Behavior Analysis Tool | 孩子学习行为分析工具 支持哪些平台？
Child Learning Behavior Analysis Tool | 孩子学习行为分析工具 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 Child Learning Behavior Analysis Tool | 孩子学习行为分析工具？
由 smyx-sunjinhui（@smyx-sunjinhui）开发并维护，当前版本 v1.0.1。
推荐 Skills