Security audit
Security checks for vulnerabilities and agentic risk
This is a Markdown-only podcast planning skill whose file and command examples are mostly expected for creating podcast project materials, with some disclosure and privacy improvements needed.
Install only if you are comfortable with a Chinese-language podcast workflow that may ask your agent to create or edit Markdown files under a podcast project directory. Review shell commands before running them, keep guest contact details minimal, and avoid storing private contact information in shared repositories or synced folders.
The display name, summary, and the entire instructional content are presented in Chinese, with no indication that users may choose another language or that the skill is intentionally limited to a Chinese-language audience. The policy calls for flagging language or locale constraints when they are imposed without user opt-in or explicit justification.
The capability table explicitly marks 'AI 生成' as unsupported in the free version. Later, the file states that an LLM API is required and that the creation flow is driven by the Agent's built-in LLM, which directly conflicts with the earlier capability claim rather than merely omitting detail.
The skill is presented as a narrow podcast-planning toolkit, but its capability section describes broad create/query/modify/delete/import/export/save behaviors driven by generic parameters. That mismatch can cause an agent or user to overtrust the scope of the skill and permit operations beyond expected podcast guidance, increasing the chance of unintended file manipulation or prompt-scope abuse when combined with read/exec tools.
The markdown includes shell commands that create directories and write files on the local filesystem, but it does not clearly warn that following or automating these steps will modify user data. In an agent setting with exec access, such examples can be operationalized automatically, leading to unintended filesystem changes or writes outside the user's intended workspace.
The skill encourages storing guest contact details and profile notes in plaintext markdown without any privacy, consent, retention, or access-control guidance. This creates unnecessary exposure of personal data, especially in shared repos, synced folders, or agent-accessible workspaces.
No suspicious patterns detected.
