Pet Body Condition & Health Analysis Skill | 宠物体态健康分析技能

Identifies obesity, emaciation, external injuries, skin abnormalities, and abnormal mental states, helping pet owners detect health issues promptly. | 宠物体态健康...

This skill does what it says (upload images/video to a backend API and return structured pet-health reports), but note the following before installing or using it: - Network uploads: user media (images/videos) and the provided 'open-id' will be sent to the configured remote API (default config points to external lifeemergence domains). If you need privacy, do not upload sensitive images or identifiers to that backend. - Local files created: the code will create/read configuration files (config.yaml) and may create a local SQLite DB under the workspace/data directory. The SKILL.md forbids reading local memory, but the implementation still initializes and uses local config/DB — this mismatch is important. - Credentials: API keys are optional but can be read from config files. If you place secrets in those config files be aware they are used for remote API calls. - No installer is provided, but the repository includes many dependency pins; running it in your environment may pull in many packages. What to do next: if you plan to use this skill, 1) confirm the target backend URL and who operates it; 2) review and, if desired, edit the config.yaml files to point to your own trusted API or remove/blank any third-party endpoints; 3) run the skill in an isolated environment (e.g., container) if you want to limit its ability to create files or access host data; 4) if you need the SKILL.md promise (never read local memory), ask the author to remove/simplify the local DB/config creation or provide proof that the code never touches local memory storage. If any of these behaviors are unacceptable, do not install or run the skill.

Type: OpenClaw Skill Name: smyx-pet-body-health-analysis Version: 1.0.0 The skill bundle contains high-risk instructions and capabilities that override standard agent behavior. Specifically, SKILL.md uses 'Mandatory Rules' to force the AI agent to bypass its local memory and long-term storage (LanceDB), mandating the use of a specific cloud API (lifeemergence.com) for all data. Additionally, scripts/smyx_common/scripts/skill.py contains logic to execute system-level commands via subprocess.run to call the 'openclaw' agent recursively. While these features appear intended for a cloud-synchronized service, the aggressive prompt-injection-style instructions to ignore local context and the automated collection of user identifiers (phone numbers/usernames) for remote authentication represent a significant deviation from standard safe skill behavior.