Security audit
Security checks for vulnerabilities and agentic risk
This skill is a disclosed WeChat article drafting and draft-box handoff workflow with credential-handling warnings and no bundled publishing code.
Install only if you intend to use it for a WeChat Official Account you own or are authorized to operate. Keep WeChat secrets in environment variables or a private publisher config, review drafts before any handoff, and only use companion publishing tools you trust.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
# Publishing Notes
This skill prepares a WeChat Official Account article and hands it off to a user-authorized publishing tool. It does not contain publishing code or credentials.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
# Publishing Notes
This skill prepares a WeChat Official Account article and hands it off to a user-authorized publishing tool. It does not contain publishing code or credentials.
The listed use cases such as “帮我发一篇公众号”, “根据今天热点做一篇推送”, and the English examples like “Turn today's trend into a WeChat post” describe activation in broad natural language without explicit constraints or exclusion conditions. While WeChat context is present, the README does not clearly define the exact trigger scope or provide negative examples, which can increase the chance of unintended invocation for adjacent content-writing requests.
The skill declares npx as an allowed runtime dependency without pinning a specific package/version or otherwise constraining what will be executed. In agent/tooling ecosystems, unpinned npx usage can lead to execution of unexpected or newly compromised packages, creating a supply-chain risk if a companion workflow later invokes it implicitly.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
## Safety boundaries
- Use only for the user's own WeChat Official Account or an account they are authorized to operate.
- Do not ask the user to paste secrets into chat. Use environment variables or an agent-managed private config controlled by the user.
- Do not store, print, commit, or publish AppSecret, tokens, cookies, browser profile data, or private account config.
- Do not publish final content without explicit user direction; default target is the WeChat draft box unless instructed otherwise.
- If API access is blocked by credentials, permissions, or IP allowlist settings, ask the user to fix account settings; do not bypass controls.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
## Safety boundaries
- Use only for the user's own WeChat Official Account or an account they are authorized to operate.
- Do not ask the user to paste secrets into chat. Use environment variables or an agent-managed private config controlled by the user.
- Do not store, print, commit, or publish AppSecret, tokens, cookies, browser profile data, or private account config.
- Do not publish final content without explicit user direction; default target is the WeChat draft box unless instructed otherwise.
- If API access is blocked by credentials, permissions, or IP allowlist settings, ask the user to fix account settings; do not bypass controls.
The README states that the skill writes Chinese WeChat articles, and the English section similarly says it drafts Chinese WeChat-style articles. This imposes a language choice by default without stating that the user can choose another language or explicitly opt into Chinese output.
This markdown file contains editorial guidance expressed through Chinese-only example headlines, which can imply a fixed language/locale requirement for generated content. Because the file does not state that the skill is intentionally China/Chinese-market specific or give the user an opt-in choice, it may violate the language/locale policy criterion.
No suspicious patterns detected.