专业宠物(猫、狗及异宠)多轮医疗问诊

Security checks across malware telemetry and agentic risk

Overview

This skill appears to do what it claims: it sends pet health questions and optional images to the VetMew API using the user's VetMew credentials, with no evidence of hidden exfiltration or destructive behavior.

Install only if you trust VetMew with the pet health information, images, and API token you provide. Use it for guidance, but seek in-person veterinary care for emergencies or serious symptoms.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

#
ASI03: Identity and Privilege Abuse
Low
What this means

The skill can act against the user's VetMew API account and may consume account quota or fail if the credential is wrong.

Why it was flagged

The skill requires a VetMew API key and secret. This credential use is expected for the stated VetMew integration, and the code uses it to sign requests to the VetMew API.

Skill content
本技能要求注入环境变量 `VETMEW_AUTH_TOKEN` (格式为 `API_KEY:API_SECRET`)
Recommendation

Use a dedicated VetMew API token if possible, avoid sharing the token in chat, and rotate it if it is exposed.

#
ASI07: Insecure Inter-Agent Communication
Low
What this means

Pet names, symptoms, health context, and optional image data or image URLs may be sent to VetMew for analysis.

Why it was flagged

The client sends signed requests to VetMew's external API endpoint. This is disclosed and purpose-aligned, but user-provided pet health details and optional images leave the local environment.

Skill content
self.base_url = "https://platformx.vetmew.com"
Recommendation

Only provide information and images you are comfortable sending to VetMew, and avoid including unrelated personal or sensitive data.

#
ASI06: Memory and Context Poisoning
Low
What this means

Future calls may continue the same VetMew conversation, which can include earlier pet-health context.

Why it was flagged

The skill instructs the agent to persist conversation IDs for reuse across turns. The persistence is scoped by consultation type, but it still preserves access to prior provider-side conversation context.

Skill content
Agent **必须** 维护独立的 Session 槽位以隔离不同类型的会话。请将 ID 持久化到对应的平台变量中
Recommendation

Reset the relevant session variable when switching pets, users, or consultation topics.

#
ASI04: Agentic Supply Chain Vulnerabilities
Low
What this means

A future install could receive newer dependency versions than those originally tested.

Why it was flagged

The Python dependencies are not version-pinned. They are common packages and proportionate to the API-client purpose, but unpinned dependencies can change over time.

Skill content
requests
python-dotenv
Recommendation

Prefer installing in an isolated environment and consider pinning reviewed dependency versions for reproducibility.