Security audit
Security checks for vulnerabilities and agentic risk
This skill is a local Chinese accounting reconciliation checker that reads user-selected materials and prints scoped findings without network use, file writes, or persistence.
Install only if you are comfortable running a local Node.js checker over the selected financial folders. Review the reported checks_not_run and sub_checks_not_run sections carefully, and treat results as reconciliation assistance rather than audit, assurance, or accounting judgment.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
This finding is a weaker variant of the same scope-mismatch issue: the documentation emphasizes one-line per-client conclusions, while the actual output may include richer structured data and partial coverage states. That is still a real trust issue, but less severe than the claims that entire check families are implemented when they are not.
Referenced artifact was not completely inspected
node scripts/run.mjs --sample # 单客户样例（内置）
Referenced artifact was not completely inspected
node scripts/run.mjs --sample # 单客户样例（内置）
Referenced artifact was not completely inspected
node scripts/run.mjs --sample # 单客户样例（内置）
Referenced artifact was not completely inspected
node scripts/run.mjs --sample # 单客户样例（内置）
The skill metadata advertises a 14-item receivables/reconciliation pack, but this file explicitly implements only the advance-receipt / revenue-recognition subset and marks other checks as withheld or out of scope. In a financial review workflow, that mismatch can mislead users into believing the full control set ran, causing material omissions in audit or risk detection even though the code itself is not overtly malicious.
The file’s user-facing natural-language strings, comments, sample input, error messages, and disclaimer are entirely in Chinese, with no indication that another language is supported or that Chinese is a documented locale-specific requirement. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
All comments, prompts, error messages, and result text presented to users are written in Chinese, including required input guidance and returned findings text. The file does not indicate that the skill is China-specific or provide any user opt-in or alternative locale, which can violate language/locale policy requirements.
This code’s natural-language interface, messages, and disclaimers are written exclusively in Chinese, and there is no indication that users may choose another language. Under the policy for natural-language violations, forcing a specific language without user opt-in should be flagged.
This JavaScript file’s natural-language interface is entirely in Chinese, including the top-level description and later user-facing advice/error text, with no indication that the skill is region-specific or that users may choose another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.
The file explicitly states it is only a free-tier subset and omits several checks, while the surrounding skill metadata promises a much broader 14-item, per-customer review. This creates a security/integrity gap: users may rely on the tool for comprehensive financial review when it actually performs only limited arithmetic checks, causing material issues to be missed without obvious warning at the skill level.