Security audit
Security checks for vulnerabilities and agentic risk
This is a mostly Chinese-language LLMOps asset-management playbook with a small local toolkit and no evidence of hidden network, credential, persistence, or destructive behavior.
Install this if you want a Chinese-language LLMOps asset-management playbook and local checklist generator. Be aware that most docs and CLI output are Chinese, the security audit contains some stale copy-paste references to a different toolkit name, and running verification scripts can overwrite files under the package's own verify directory.
A declared LLMOps asset-management skill that actually performs unrelated security/quality radar chart generation is a significant transparency and trust violation. Users may execute or rely on tooling under false assumptions, which can conceal unexpected data access, misleading outputs, or repurposed code paths not justified by the stated function.
This markdown file presents all user-facing audit content exclusively in Chinese, which effectively forces a specific language/locale on readers. The policy allows locale constraints only when the skill offers user choice or clearly documents a justified region-specific need, neither of which appears here.
The skill advertises executable local tooling but does not declare any explicit tool scope or permissions, leaving users and hosting platforms without a clear boundary on file, shell, or possible network behavior. Even if the markdown only documents commands, missing permission declarations weakens least-privilege controls and can enable unexpected access if the referenced tooling is executed.
The main instructional content, usage guidance, and FAQ are presented in Chinese, while only the metadata field description_en provides English text. There is no statement that users may choose their preferred language, so the file imposes a locale/language preference in natural-language content without explicit opt-in.
This markdown file presents all guidance in Chinese, including the title and all FAQ entries, but does not indicate that the skill is region-specific or that users may choose another language. Under the policy for natural-language violations, forcing a specific language without user opt-in is a reportable issue.
This code file contains natural-language strings that force a specific language/locale for all users, including the module docstring and CLI descriptions. Under the policy, language constraints should either be optional for the user or clearly documented and justified as region-specific.
The file’s natural-language content, generated SVG labels, and console output are all hard-coded in Chinese, with no option for users to select another language or any justification that the skill is intended only for a Chinese locale. This matches the policy concern for forcing a specific language without user opt-in.
The module docstring is entirely in Chinese and frames the skill behavior and evaluation dimensions only in that language. Under the policy, forcing a specific language without opt-in is a natural-language locale violation unless the constraint is explicitly justified, which is not present here.
subprocess module calls execute external commands. Without careful input validation, this enables command injection.
def run_tool(args, timeout=30):
try:
r = subprocess.run([PY, os.path.join(TOOLS, "llmops_toolkit.py")] + args,
capture_output=True, text=True, timeout=timeout)
return r.returncode, r.stdout or ""
except Exception as e:
Line L04 requires branding to use a specific Chinese name form exclusively ("统一用「注册老炮」") and forbids alternatives, but the document does not provide any user choice or clear locale-specific justification. This is a natural-language policy concern because it imposes a fixed language/locale presentation requirement rather than offering opt-in flexibility.
This manifest contains natural-language metadata in Chinese (package name and author) but does not indicate whether the skill is intended to be Chinese-only or whether users can choose another language. Under the policy rule for language/locale constraints, forcing or implying a single language without opt-in can be a policy concern.
SQP-3 applies to all file types and covers natural-language policy issues such as forcing a specific language without user opt-in. This reference material presents all guidance only in Chinese and does not mention that the language is optional, selectable, or limited to a justified locale-specific context.
This SVG contains all user-visible labels and captions in Chinese, including the title, axis labels, legend, and footer text. For a generally reusable skill asset, forcing a specific language without opt-in or documented regional scope can violate language/locale policy expectations.
This JSON file uses Chinese natural-language labels and notes throughout, such as the note field and many check names, but does not indicate that the skill is region-specific or that users may choose a language. The stated policy requires flagging language or locale constraints when they are imposed without opt-in or clear justification.