Security audit
Security checks for vulnerabilities and agentic risk
The skill is a local summarization tool, but it keeps submitted content in a persistent searchable database and can export plaintext files without clear user controls.
Install only if you are comfortable with summaries and original submitted content being stored locally in ~/.ai_summary.db and Markdown exports being written as plaintext files. Avoid using it for confidential meeting notes or sensitive project data unless you manage the database and exported files yourself.
The skill advertises no explicit tool scope, yet its documented capabilities imply file writing and possibly shell-backed execution via the referenced script and Markdown export flow. Without declared restrictions, an agent may grant broader-than-necessary access, increasing the chance of unintended file modification or misuse if the implementation is later expanded or invoked in unsafe contexts.
The manifest description and the entire skill documentation are presented only in Chinese, with no indication that users may choose another language or that the skill is intentionally limited to a Chinese-only context. This can violate language or locale policy where skills should not force a specific language without opt-in or documented justification.
The invocation examples use very generic natural-language phrases like '总结今天的日报' and '复盘这个项目', which can cause accidental triggering during ordinary conversation. Unintentional activation is risky here because the skill stores content locally and may process sensitive work notes or project data without the user realizing the skill was invoked.
The skill mentions a local database path, but the user-facing description and examples do not prominently warn that summaries and reviews are persisted locally. This creates a privacy risk because users may provide meeting notes, project retrospectives, or daily reports containing sensitive information without informed consent about storage.
The skill presents itself as a summarization/review tool but also persists full user content and generated summaries to a local SQLite database and supports file export. This creates a confidentiality risk because users may submit sensitive notes, meeting content, or project retrospectives without realizing the data will be stored long-term on disk.
The code keeps a searchable history of prior user submissions, which expands the privacy exposure beyond one-time summarization into long-term retention and retrieval of potentially sensitive text. If another local user or process can access the database, past notes and summaries can be enumerated and searched easily.
At this point the program automatically saves raw content and summary metadata to a local database when a title is provided, without an explicit warning or consent prompt about persistent storage. Because summarized content often includes sensitive business or personal information, silent local persistence can lead to accidental data exposure on shared or monitored systems.
The export function writes markdown to disk and defaults to a predictable filename without warning the user about file creation or possible overwrite/placement risks. This can leak sensitive summary data into the current working directory, backups, sync folders, or shared locations unintentionally.
The file comments are written entirely in Chinese, indicating the skill context may presume a specific language/locale. The policy requires flagging language constraints when a skill appears to force a specific language without offering user choice or documenting the constraint.
Natural-language strings in the module description and CLI are Chinese-only, which effectively forces a specific language experience on users. The file does not indicate that the tool is region-specific or provide any opt-in or alternative locale handling.
No suspicious patterns detected.
