Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Alphapai Research

v1.0.0

Alpha派金融投研平台API技能,用于调用Alpha派(AlphaPai/PaiPai)的投资研究接口。覆盖五大核心能力:投研知识问答、投研数据检索、投资研究Agent(公司一页纸/业绩点评/调研大纲/主题选股/投资逻辑/可比公司/观点Challenge/行业一页纸/个股选基/主题选基/画图)、股票公告列表查询...

0· 68·2 current·2 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The skill's stated purpose (call AlphaPai 投研 APIs) matches the included client code and CLI instructions. However, the package ships with a prefilled config.json containing an API key and base_url which effectively grants access to a remote AlphaPai service without requiring the user to provide credentials. The registry metadata declares no required credentials/env vars, but a credential is embedded in the bundle — this is an incoherence and unexpected for a skill that declares no secrets.
Instruction Scope
SKILL.md confines the agent to using the provided CLI wrapper (scripts/alphapai_client.py) and instructs strict behavior for presenting API responses (present raw output without summarizing). That is consistent with the stated purpose. The instructions also instruct the agent to always use the skill when certain keywords appear, which will cause outbound API calls automatically; this is logical for the feature but increases the chance of sending user data to the external service.
Install Mechanism
No install spec — instruction-only plus an included Python script. No downloads/third‑party install steps are declared, so nothing will be pulled at install time. Risk comes from runtime network requests performed by the included script rather than from installation.
!
Credentials
No environment variables or credentials are declared by the skill, yet config.json (bundled with the skill) contains a clear-text API key ("niili8cng7x3azxbfxaxgp2m") and base_url (https://open-api.rabyte.cn). The client will use that key by default. This grants outbound access under a third party's account and is disproportionate/unexpected given the declared requirements. The client also hardcodes a userInfo block (userName: "rabyte"), indicating queries may be associated with that account.
!
Persistence & Privilege
The skill does not request always:true and is user-invocable, which is normal. However, the embedded config.json makes a persistent credential part of the skill package (persistent presence of a usable API key in the skill files). The client also writes to config.json when the user runs the provided config command — that write behavior is normal, but bundling an active key is a persistent privilege and a privacy/authentication concern.
What to consider before installing
This skill will send your questions and context to an external AlphaPai endpoint. Important points to consider before installing: - A usable API key (and base_url) is bundled in config.json inside the skill (api_key: "niili8cng7x3azxbfxaxgp2m"). That means the skill will call the remote service under the packager's account by default — your queries and any data sent could be logged by that third party. - The skill declares no required credentials, but includes a credential in the package (incoherent). If you care about who is billed/has access to your query data, replace or remove config.json and set your own API key with the CLI (python scripts/alphapai_client.py config --set-key YOUR_KEY) before using. - The SKILL.md requires presenting API responses verbatim (no summarization). This can cause long, unfiltered outputs (including potentially sensitive or copyrighted content) to be shown to users exactly as returned by the remote service. - There is no homepage or source contact; the package owner is unknown. If you need accountability for data handling, request the publisher's identity, confirm the legitimacy of the embedded key, and review the remote service's privacy/TOS. Recommended actions: 1) Do not use the bundled key. Delete or rename config.json in the skill bundle, then run the provided config command to set your own key if you have one. 2) If you must use the skill, obtain an API key you control (rotate/create a dedicated key), and verify the service's data retention/privacy policy before sending sensitive data. 3) If you cannot verify the origin of the embedded key or the vendor, avoid installing or use the skill in an isolated environment and treat all outputs as coming from a third party. If you want, I can list the exact files and lines that reveal the embedded key and the network endpoints to help you remove or modify them safely.

Like a lobster shell, security has layers — review code before you run it.

latestvk971xa8atx7ygjmxm8417w5csd83cp56

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments