Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent Chinese-language enterprise alias generator, with offline defaults and optional public web lookups that users should avoid using for confidential company names.
Install only if a Chinese-language enterprise alias workflow fits your use case. Keep the default offline mode for private or internal company names; enable or run the web-search modules only when you are comfortable sending the queried company name to public third-party services.
The trigger phrases are broad and include common requests such as '提取关键字' and general alias-generation language, which can cause the skill to activate during ordinary conversation unrelated to an explicit request to use this skill. Over-broad activation increases the chance of unintended interception of user queries and incorrect routing, especially in systems with multiple skills.
The manifest presents the skill name, description, and tags only in Chinese, with no indication that the skill is China-specific or that users can opt into another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy concern.
This file defines all entity names and aliases exclusively in Chinese string literals, with no accompanying comment, documentation, or option indicating that the skill is Chinese-only. Under the policy, forcing a specific language or locale without user opt-in is a natural-language policy concern.
This code forces use of Chinese-language data sources by setting the Wikidata search language to Chinese and targeting zh.wikipedia.org, which imposes a locale choice on users. The file contains no opt-in, configuration, or justification that this skill is intended only for a Chinese-language or China-specific context.
Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.
"props": "claims",
"format": "json"
}
response = requests.get(self.wikidata_url, params=entity_params, timeout=10)
if response.status_code != 200:
return []
Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.
"format": "json",
"utf8": ""
}
response = requests.get(self.wikipedia_url, params=content_params, timeout=10)
if response.status_code != 200:
return []
This Python file contains natural-language documentation and CLI text entirely in Chinese, including the module docstring and argument descriptions. Because the skill does not offer a user language/locale choice or state that it is intentionally restricted to a Chinese-language context, it creates a language policy concern under the natural-language policy rule.
The generate_aliases docstring states use_web means web search and labels it as '暂未实现' (not yet implemented) at L365, but the module already contains a working search_wiki function that performs HTTP requests to Wikipedia when USE_WIKI is enabled (L263-L310), and rule-based generation invokes it at L329-L335. This is an active contradiction between the code documentation and actual implemented behavior.
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
def fetch_url(url: str, timeout: int = 10) -> Optional[str]:
"""使用 curl 获取网页内容"""
try:
result = subprocess.run(
['curl', '-s', '-A', 'Mozilla/5.0',
'--connect-timeout', str(timeout),
'--max-time', str(timeout * 2),
The skill sends the user-provided company name to Baidu, Qichacha, and other third-party sites without notice or consent. If the queried company name is sensitive, internal, or tied to confidential research, this leaks user intent and data to external services and may create compliance or privacy issues.
The changelog content switches into Chinese for section headings and descriptions, which imposes a specific language on readers. There is no indication that the skill is region-specific or that users can choose an alternate language, matching the language/locale policy concern.
All user-facing instructions, triggers, examples, and behavior descriptions are presented only in Chinese, and the document does not indicate that the user can choose another language. Under the policy, forcing a specific language without opt-in can be a locale-policy issue unless the restriction is explicitly justified.
Natural-language content in the module docstring and runtime messages is exclusively Chinese, and the script offers no language selection or opt-in. Under the policy, forcing a specific language without user choice is a language/locale policy violation unless clearly justified as region-specific.
This code file presents its descriptive natural-language content entirely in Chinese, including the header comments and function docstring, without indicating that the skill is China-specific or offering any language/locale choice. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation.
At L204 the comment says '精确匹配' (exact match), but L206 uses if key in company_name, which is substring matching rather than equality. At L209 the comment says fuzzy matching by taking the first 3 characters, but L212 again uses broad containment checks, so the documentation overstates the precision and nature of the matching behavior.
No suspicious patterns detected.