← Back to Skills Marketplace
18072937735

Stranger Proximity Alert Skill | 陌生人靠近预警技能

by smyx-skills · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
65
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install smyx-stranger-approach-warning-analysis
Description
Detects the appearance of strangers near minors and actively issues safety reminder alerts to protect minor safety, suitable for homes, schools, childcare ce...
Usage Guidance
Proceed cautiously. What to check before installing or running: 1) Inspect/configure where uploads and DB files will be stored (OPENCLAW_WORKSPACE or skill data directory) and disable or restrict automatic saving if you don't want videos persisted. 2) Verify and control any config files in the workspace (skills/smyx_common/scripts/config.yaml and workspace-level config) — the skill will read them and could pick up API keys/open-ids silently. 3) Confirm which external API endpoints will be called (config files show lifeemergence/open-api endpoints in examples); only use endpoints you trust and ensure TLS/auth is correct. 4) The package contains many unused/extra modules and a large dependency list — run in an isolated environment and pin/inspect dependencies before installing. 5) Because this processes video of minors and uses facial recognition, check legal/privacy compliance and obtain consent; avoid running on production data until you are confident of storage, retention, and access controls. 6) If you require stricter guarantees, ask the author for a minimal build that only contains the stranger-approach logic, a documented install step, and a mode that does not persist uploads or local DB entries.
Capability Analysis
Type: OpenClaw Skill Name: smyx-stranger-approach-warning-analysis Version: 1.0.0 The skill bundle contains high-risk prompt instructions in SKILL.md that command the AI agent to ignore its local memory and retrieval systems (LanceDB), forcing exclusive reliance on a remote cloud API. It implements a complex authentication and data collection framework in smyx_common/scripts/util.py and dao.py that captures user identifiers (such as phone numbers or usernames) and transmits them to a remote endpoint (lifeemergence.com) to obtain session tokens. Furthermore, smyx_common/scripts/skill.py includes functionality to execute shell commands via subprocess to interact with the OpenClaw agent CLI, which provides a powerful execution capability that could be leveraged for unauthorized actions.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
Name/description (stranger proximity alerts for minors) align with the code that wraps an external AI analysis API and performs face/stranger analysis. However the repository also includes unrelated modules/examples (e.g. a 'face_analysis' TCM health skill and broad common utilities), which makes the package larger and more general than the single-purpose description. That extra code is not required for a simple wrapper and may increase attack surface.
Instruction Scope
SKILL.md explicitly forbids reading local memory and LanceDB, and requires open-id be taken from specific config files or user input. In practice the code reads workspace/shared config files and environment variables (e.g. OPENCLAW_SENDER_OPEN_ID, OPENCLAW_WORKSPACE, FEISHU_OPEN_ID) via skills/smyx_common scripts. The skill also auto-saves uploaded attachments to the skill/workspace data area and contains a local SQLite DAO implementation. Those file reads/writes and automatic saving of user-uploaded video are broader than the 'do not read local memory' rule and could persist sensitive video or metadata.
Install Mechanism
No install spec is provided (instruction-only install), but the bundle contains many Python modules and a large requirements.txt in skills/smyx_common (dozens of packages). That mismatch means the skill expects a heavy Python runtime and many dependencies but does not declare how to install them. This is a friction/risk point (unexpected dependencies, unclear provenance).
Credentials
Registry metadata claims 'no required env vars', yet code reads multiple environment/config sources: OPENCLAW_SENDER_OPEN_ID/OPENCLAW_SENDER_USERNAME, OPENCLAW_WORKSPACE, and will load API config from skills/smyx_common/scripts/config.yaml or workspace-level config. The open-id acquisition flow reads config files in the skill directory and in a shared workspace; that can cause the skill to silently pick up workspace-level API keys or IDs. The CLR between declared and actual env/config access is inconsistent and meaningful for privacy/security.
Persistence & Privilege
The skill will (per SKILL.md) automatically save uploaded attachments into an attachments directory and the code includes a local SQLite Dao that writes to workspace/data (smyx-common-claw.db). Even though always:false and not forced into every agent run, the skill persists sensitive media and local DB files to the workspace without an explicit consent step documented in the metadata. That persistent storage of video of minors is sensitive and should be reviewed/controlled.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install smyx-stranger-approach-warning-analysis
  3. After installation, invoke the skill by name or use /smyx-stranger-approach-warning-analysis
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release – Stranger Approach Warning Analysis: - Detects strangers approaching minors using video/image analysis and issues real-time safety alerts. - Supports customizable safety zones, personnel whitelists, and multi-level risk warnings for homes, schools, and care centers. - Enforces cloud-based history reporting; strictly prohibits reading/writing local memory for reports. - Requires strict open-id acquisition flow (config file/user prompt); cannot auto-generate or assume open-id values. - Provides comprehensive CLI usage instructions and outputs standardized Markdown table for history queries.
Metadata
Slug smyx-stranger-approach-warning-analysis
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Stranger Proximity Alert Skill | 陌生人靠近预警技能?

Detects the appearance of strangers near minors and actively issues safety reminder alerts to protect minor safety, suitable for homes, schools, childcare ce... It is an AI Agent Skill for Claude Code / OpenClaw, with 65 downloads so far.

How do I install Stranger Proximity Alert Skill | 陌生人靠近预警技能?

Run "/install smyx-stranger-approach-warning-analysis" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Stranger Proximity Alert Skill | 陌生人靠近预警技能 free?

Yes, Stranger Proximity Alert Skill | 陌生人靠近预警技能 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Stranger Proximity Alert Skill | 陌生人靠近预警技能 support?

Stranger Proximity Alert Skill | 陌生人靠近预警技能 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Stranger Proximity Alert Skill | 陌生人靠近预警技能?

It is built and maintained by smyx-skills (@18072937735); the current version is v1.0.0.

💬 Comments