Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
向日葵远程控制
v1.0.0向日葵远程控制(awesun-remote-control) 提供 22 个工具。使用场景包括:控制命令、控制连接、控制断开。关键词:远程控制,设备管理,桌面控制,远程CMD,远程电源管理。
⭐ 0· 278·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The name, README, SKILL.md and executor.py all describe an Awesun MCP-based remote-control skill and the provided files implement a local MCP client that launches the MCP server binary and talks to it via stdio. However the registry metadata lists no required binaries or config paths while the README and mcp-config.json clearly require the Awesun MCP server binary and an API token — a mismatch between declared requirements and what the skill actually needs.
Instruction Scope
SKILL.md and executor.py confine runtime actions to interacting with an MCP server (listing tools, describing tools, calling tools). They do not instruct reading arbitrary system files or harvesting unrelated environment variables; runtime behavior is limited to spawning the configured MCP server command and exchanging MCP protocol messages.
Install Mechanism
There is no installer; this is instruction/code-only. package.json includes a setup script invoking 'pip install mcp' and executor.py requires the 'mcp' Python package. This is a moderate-risk pattern because it relies on a PyPI package (mcp) and launching a local binary specified by the user — but there is no remote download-from-arbitrary-URL or opaque installer in the skill itself.
Credentials
The skill does not declare required env vars in its registry metadata, but mcp-config.json (bundled) contains an AWESUN_API_TOKEN placeholder and AWESUN_API_URL; the skill needs that token/URL to control devices. Requesting a local MCP token is proportionate to a remote-control skill, but the token is stored/used via a config file rather than declared environment requirements — this storage/usage of a sensitive token should be understood by the user.
Persistence & Privilege
The skill is not set to always:true and does not request persistent system-level privileges. It executes the configured MCP server process and communicates with it, but it does not appear to modify other skills or system-wide configuration beyond its own config file.
What to consider before installing
What to consider before installing:
- This skill implements an Awesun (Awesun-MCP) client: it launches a local MCP server binary (path is taken from mcp-config.json) and uses a token (AWESUN_API_TOKEN) to communicate. That token grants the skill the ability to control remote machines — only provide it if you trust the target MCP server and network.
- The package metadata did not declare the runtime dependency on the Awesun MCP binary or the fact that you must edit mcp-config.json; verify and edit mcp-config.json yourself so the command path points to the expected local binary and AWESUN_API_URL is localhost (127.0.0.1) if you want to limit exposure.
- The skill requires the 'mcp' Python package (pip). If you plan to run it, inspect the 'mcp' package source and the awesun-mcp project on GitHub to ensure they are the legitimate projects you expect.
- This skill exposes powerful capabilities (remote command execution, remote desktop, screenshots, wake/shutdown). Only enable or use it in trusted environments and with accounts/tokens that have minimal necessary privileges. Be cautious about autonomous invocation: although the skill is not 'always:true', skills can be invoked by an agent; avoid enabling it in agents or contexts that have the authority to act on your behalf without supervision.
- If you need higher confidence: ask the author for provenance of the mcp package and awesun-mcp binary you will run, or run the executor in a sandboxed environment first and monitor network/process behavior.mcp-config.json:5
Install source points to URL shortener or raw IP.
About static analysis
These patterns were detected by automated regex scanning. They may be normal for skills that integrate with external APIs. Check the VirusTotal and OpenClaw results above for context-aware analysis.Like a lobster shell, security has layers — review code before you run it.
latestvk977ymwg3vf92nbc22hrgc7y7982qdkw
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
