← 返回 Skills 市场
Visual Emotion Recognition Skill | 人体视觉情绪识别技能
作者
smyx-skills
· GitHub ↗
· v1.0.0
· MIT-0
70
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install smyx-human-emotion-recognition-analysis
功能描述
Uses visual AI on frontal faces to recognize multi-dimensional emotions like happiness, sadness, depression, calmness, anger, surprise, and fear in real-time...
安全使用建议
Key things to consider before installing or running this skill:
- Data privacy: The skill processes face images/videos. Do not upload sensitive or identifying material unless you trust the remote API endpoints and have consent from subjects.
- Undeclared environment and file access: Although the skill metadata lists no required env vars, the code reads OPENCLAW_SENDER_OPEN_ID, OPENCLAW_WORKSPACE, FEISHU_OPEN_ID and will look for config.yaml files in the skill and shared smyx_common directory for api-key/base URLs. Confirm what values those files/vars contain in your environment before use.
- Local persistence: The bundle includes a DAO that creates a local SQLite DB under a workspace/data directory and the skill will save attachments to disk. If you need no local traces, do not run this skill without modifying it.
- Inconsistent promises: The SKILL.md forbids reading local memory yet the codebase includes local storage utilities and DB logic; ask the author to clarify and to provide a minimal, audited implementation that strictly follows the stated privacy rules.
- Dependencies and installation: A large requirements.txt is present, but no install script is provided. Running the scripts may require installing many Python packages—review those packages (smyx_common/requirements.txt) and install them in an isolated environment if you proceed.
- Endpoint and API keys: The repo uses base URLs (e.g., lifeemergence.com / open.lifeemergence.com) defined in config files. Verify those endpoints and never hardcode or expose real API keys or production user identifiers to untrusted third parties.
Suggested next steps: ask the publisher for (1) source/origin and intended endpoint, (2) a clear privacy statement and whether local DB usage is intentional, (3) a trimmed minimal version that only includes the emotion-recognition paths you need, or (4) run the code in an isolated sandbox and inspect the files it writes before giving it real data.
功能分析
Type: OpenClaw Skill
Name: smyx-human-emotion-recognition-analysis
Version: 1.0.0
The skill bundle provides human emotion recognition and TCM face analysis by interfacing with external APIs at lifeemergence.com. It features a sophisticated common utility layer (smyx_common) that manages local SQLite storage for user tokens and includes capabilities for recursive agent calls via the OpenClaw CLI. While SKILL.md contains 'Mandatory Memory Rules' that instruct the AI agent to ignore local memory files and prioritize cloud data, these appear to be functional design choices to ensure data consistency with the remote backend rather than malicious prompt injections. The code follows standard practices for API integration, including file validation and structured error handling.
能力标签
能力评估
Purpose & Capability
The code and SKILL.md align with a cloud-backed facial emotion recognition service: scripts call a remote API, accept local files or URLs, format reports, and support listing historical reports. However, the bundle also contains reused/leftover code and docs for pet/health/TCM face-analysis, indicating copy-paste/generalization across domains (e.g., 'pet' mentions and health endpoints). That is not fatal but reduces confidence that all requested behaviors are limited to the stated human-emotion use-case.
Instruction Scope
The runtime instructions mandate checking config files in two filesystem locations and require an open-id before calling the API. The skill saves uploaded attachments to an attachments directory and instructs running local scripts (python -m scripts.human_emotion_recognition_analysis). The SKILL.md explicitly forbids reading local memory files, yet the codebase contains local DB/DAO modules and logic that can write/read local data under a workspace 'data' directory. The instructions also require executing scripts from the skill root. These behaviors expand file and persistence access beyond a simple transient API call.
Install Mechanism
No install spec is provided (instruction-only), but the repository includes many Python modules and a large requirements.txt under skills/smyx_common. That mismatch means the skill expects a Python environment with many dependencies but provides no automated install procedure—an operational inconsistency (not necessarily malicious) that can cause unexpected failure or require the agent/operator to install packages manually.
Credentials
Registry metadata declares no required env vars, but the code reads environment variables and workspace paths (e.g., OPENCLAW_SENDER_OPEN_ID, OPENCLAW_SENDER_USERNAME, OPENCLAW_WORKSPACE, FEISHU_OPEN_ID) and also reads config.yaml files inside the skill and a shared skills/smyx_common path for api-key/base URLs. The SKILL.md lists an explicit open-id retrieval order that mentions config files and user input but does not document using these environment variables—this is an undeclared credential/config access pattern and therefore disproportionate to the published metadata.
Persistence & Privilege
Although 'always' is false and the skill is not forced into every agent run, the code contains a local DAO that creates/uses a SQLite DB under a workspace/data path and the skill will save uploaded attachments to disk. That means the skill creates persistent artifacts on disk (database files and attachments) even though the SKILL.md attempts to require cloud-only history queries. This creates a mismatch between stated 'no local memory usage' policy and the actual ability to persist data locally.
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-human-emotion-recognition-analysis - 安装完成后，直接呼叫该 Skill 的名称或使用
/smyx-human-emotion-recognition-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
v1.0.0
Initial release of the human-emotion-recognition-analysis skill.
- Real-time, multi-dimensional emotion recognition from frontal face images or videos, supporting happiness, sadness, depression, calmness, anger, surprise, and fear.
- Quantifies emotion intensity and marks abnormal emotional states; suitable for human-computer interaction and mental health monitoring.
- Enforces strict open-id acquisition for user identification, with mandatory cloud-based retrieval of historical emotion recognition reports (no local memory usage allowed).
- Automatically saves uploaded media files and provides structured emotion analysis reports, with clear usage flow and sample commands.
- Outputs historical report lists as Markdown tables with direct links to full reports.
元数据
常见问题
Visual Emotion Recognition Skill | 人体视觉情绪识别技能 是什么？
Uses visual AI on frontal faces to recognize multi-dimensional emotions like happiness, sadness, depression, calmness, anger, surprise, and fear in real-time... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 70 次。
如何安装 Visual Emotion Recognition Skill | 人体视觉情绪识别技能？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install smyx-human-emotion-recognition-analysis」即可一键安装，无需额外配置。
Visual Emotion Recognition Skill | 人体视觉情绪识别技能 是免费的吗？
是的，Visual Emotion Recognition Skill | 人体视觉情绪识别技能 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Visual Emotion Recognition Skill | 人体视觉情绪识别技能 支持哪些平台？
Visual Emotion Recognition Skill | 人体视觉情绪识别技能 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 Visual Emotion Recognition Skill | 人体视觉情绪识别技能？
由 smyx-skills（@18072937735）开发并维护，当前版本 v1.0.0。
推荐 Skills