Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed Chinese financial-news search helper that sends user queries to the documented Eastmoney endpoint and saves returned results locally.
Install only if you are comfortable sending financial search queries to Eastmoney's API using your MX_APIKEY and retaining result files, including raw JSON, under the documented output directory. Avoid putting client secrets, private account details, or confidential investment research directly into search queries.
The skill declares capabilities that imply environment variable access, file writes, and outbound network use, but it does not explicitly scope or constrain those permissions. This increases the risk of over-privileged execution, making it easier for the skill or a future modification to access secrets, write arbitrary files, or transmit data beyond what is necessary.
The skill sends user-supplied query content to an external third-party API endpoint, which is an actual data exfiltration boundary even if it is expected functionality. In a financial research context, user queries may contain sensitive investment interests, proprietary analysis topics, or client-related information, so transmission to an external service carries privacy and confidentiality risk.
- 凭据保护: API Key 仅通过环境变量
MX_APIKEY在服务端或受信任的运行环境中使用，不会在前端明文暴露。
curl -X POST --location 'https://mkapi2.dfcfs.com/finskillshub/api/claw/news-search' \
--header 'Content-Type: application/json' \
--header 'apikey: YOUR_API_KEY' \
--data '{"query":"立讯精密的资讯"}'
The file’s user-facing natural-language strings, comments, usage text, and output are entirely in Chinese, and the skill description identifies it specifically as a Chinese financial news search skill. There is no indication that users may choose another language or that the Chinese-only behavior is a documented, justified locale requirement, which matches the language/locale policy violation criteria.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
"query": query
}
response = requests.post(self.BASE_URL, headers=headers, json=data, timeout=30)
response.raise_for_status()
return response.json()
The skill is described as a search capability, but it also writes retrieved content and metadata to local disk by default. In agent environments, queries and results may contain sensitive financial research topics, user interests, or proprietary prompts, so silent persistence expands the data exposure surface beyond the stated purpose.
The code saves the full raw JSON response to disk even though the skill's stated purpose is only to perform search and return results. Raw responses often contain more fields than are shown to the user, which can include extra metadata, identifiers, or provider content that increases privacy, compliance, and data handling risk if files are later accessed by other tools or users.
The natural-language content of the skill, including the manifest description and README sections, is presented only in Chinese. Under the policy, forcing a specific language without user opt-in can be a locale/language policy violation unless the constraint is explicitly justified as region-specific.
No suspicious patterns detected.