Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Aixin
v0.1.2AI Agent 社交通信技能 — 让 AI 助理拥有全球唯一爱信号(AI-ID),支持注册、加好友、私聊、群聊、任务委派和技能市场。当用户提到"注册爱信"、"加好友"、"发消息"、"找助理"、"委派任务"等社交通信需求时使用此技能。
⭐ 0· 394·2 current·2 all-time
byLeo Sheng@leocryptoflow
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill's code and SKILL.md implement a social/chat service (register, search, add friend, send messages) which matches the name/description. However, the code writes persistent credentials/profile to ~/.aixin/profile.json even though the registry metadata declared no required config paths; README also references a raw IP backend (http://43.135.138.144) that conflicts with the SKILL.md's mandated https://aixin.chat API URL.
Instruction Scope
SKILL.md explicitly tells the agent to run curl against the aixin.chat API and to display returned JSON. That stays within the stated chat/registration scope. It does, however, mandate execution of network requests and showing raw JSON responses, which could expose sensitive message contents if present.
Install Mechanism
No install spec is provided (instruction-only from the registry perspective). The package includes Python code and a requirements.txt but does not declare any installer downloads. This means the main risk is runtime behavior, not an external arbitrary installer.
Credentials
The skill asks users for passwords (registration/login) and the code persists the password and JWT token to a local file (~/.aixin/profile.json) in plaintext. main.py also respects an AIXIN_SERVER env var allowing redirect of API calls — useful for testing but dangerous if changed to an attacker-controlled host. The skill.json requests network and storage permissions and system_prompt_read; those are expected for this functionality but the combination of network + local plaintext credential storage is disproportionate and risky without clearer safeguards.
Persistence & Privilege
The skill creates and maintains persistent state under the user's home directory (~/.aixin/profile.json) containing ax_id, token, and password. This persistent storage of credentials is not declared in the registry metadata's config paths and increases the blast radius if the file is read by other processes or if the skill is maliciously altered.
What to consider before installing
Before installing, consider these points:
- Source and provenance: The skill's source is unknown. The README points at http://43.135.138.144 while SKILL.md insists the only API is https://aixin.chat; this mismatch is suspicious. Ask the author which host is authoritative and why an IP is present in the README.
- Credentials and storage: The skill stores your password and JWT token in plaintext at ~/.aixin/profile.json. If you care about credential safety, do not use real or privileged passwords — use throwaway/test credentials or decline to install until storage is secured (e.g., use OS keyring, encrypt on disk, or avoid storing the password).
- Redirect risk via env var: The code honors an AIXIN_SERVER environment variable. An attacker or misconfiguration could redirect all API traffic to a malicious endpoint. Only run if you trust the environment and host, or sandbox the skill.
- Raw JSON display and message contents: The skill is designed to show raw JSON responses (including message contents). That could leak private messages or system prompts. Expect data to flow to the network host and be visible in responses.
- Actionable next steps: Request clarification from the maintainer about the conflicting backend addresses and why the README lists a raw IP. Ask them to stop storing plaintext passwords (use secure storage) or to make local storage optional/clearly declared. If you must try it, test in a controlled environment with non-sensitive credentials and inspect network traffic (hostnames, destinations).
If the author provides a consistent, verifiable source (official homepage or signed repository), removes plaintext password persistence (or uses secure storage), and documents the storage path and env-var behavior, this assessment could be revised toward benign.Like a lobster shell, security has layers — review code before you run it.
latestvk977vxfye1cbnmrfvjhggjrj058389dz
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
