Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Xyfcli Order Guide
v1.4.4肥料订单下单与客户/供应商管理技能。使用 xyfcli CLI 执行:(1) 下单/订货/采购肥料;(2) 查询产品信息/价格/库存;(3) 查询客户/供应商/经销商信息;(4) 解析订货单图片。触发词:下单、订货、买肥料、查产品、查客户、供应商信息、产品列表、客户编码。
⭐ 0· 345·1 current·1 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
Name/description match the actual behavior: the skill drives an xyfcli command-line client to query products, customers, addresses and place orders. Declared requirement (binary 'xyfcli') is appropriate. However, the package also includes a full xyfcli Python implementation in the skill bundle but provides no install spec — this mismatch (instruction-only metadata vs included code) is unexpected and should be clarified.
Instruction Scope
SKILL.md instructs only domain-relevant actions (running xyfcli commands, OCR of order images, queries and validations). It does not instruct reading unrelated system files or unrelated credentials. The guides explicitly require confirmation before placing orders.
Install Mechanism
No install spec is declared (instruction-only), which minimizes automated installation risk. But the skill bundle contains a complete Python CLI (pyproject.toml, package source, README with manual install steps). That means a user or operator will likely run a manual install command to get the 'xyfcli' binary — the install step is not automated and is left to the user. This inconsistency increases operational risk because users may unknowingly install bundled code.
Credentials
The skill declares no required env vars and only requires the xyfcli binary. The included CLI stores API base URL and an authorization token in ~/.xyfcli/config.json; it will ask the user to set a token to communicate with the backend. Storing a token locally and allowing an arbitrary base_url is expected for this use case but means the token and base URL choice are sensitive — if an attacker or misconfigured base_url is used the CLI could send data to an arbitrary endpoint.
Persistence & Privilege
The skill does not request always:true and does not ask to modify other skills. However, the included Python code (api_client/config) will create a ~/.xyfcli directory and write a default config.json if the config is missing (side-effect at import). This is persistent filesystem activity that users should be aware of before installing/ importing the bundled CLI.
What to consider before installing
Plain-language checklist before installing or running this skill:
- The skill uses an external CLI named 'xyfcli' (declared requirement). The repository actually includes a full Python implementation of that CLI, but there is no automated install step — the README shows a manual install command. Don’t assume the CLI is already installed; installing the bundled code will run user-owned code on your machine.
- Installing or importing the bundled CLI will create a ~/.xyfcli/config.json file (the code writes a default config if missing). That file will hold the API base URL and any authorization token you set. Only set the token if you trust the backend and know which endpoint you are targeting.
- Default API base is http://127.0.0.1:8000 in the sources; if you change base_url to a remote server, the CLI will send requests (and the token) to that server. Make sure the base_url points to a trusted internal system before setting the token.
- The skill’s runtime instructions and code appear focused on legitimate ordering tasks and do not request unrelated secrets, but because the skill can talk to an API and persist a token, review the included code (api_client/config/order/shop) yourself or ask an administrator to do so before installing.
- If you want minimal risk: do not install the bundled CLI; instead verify you have an existing trusted 'xyfcli' binary from a vetted source, or ask the skill author/maintainer for an official installation package or signed release. Also, confirm whether the agent will be allowed to autonomously invoke the skill (default = permitted); if you’re concerned about accidental automated orders, restrict autonomous invocation until you’ve tested it in a safe environment.
If you want, I can list the exact files and code locations you should inspect (e.g., where the config is created and where HTTP endpoints are called) or produce a short step-by-step safe-install checklist.Like a lobster shell, security has layers — review code before you run it.
latestvk97e4tjm9ctwhqrsr8x3adcqen841jqj
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
Binsxyfcli
