Security audit
Security checks for vulnerabilities and agentic risk
This is a Chinese-language Feishu help guide with some privacy and data-loss cautions missing, but it does not install code, run commands, or take actions on the user's behalf.
Reasonable to install as a Feishu reference skill, especially for Chinese-speaking users. Treat instructions involving public sharing links, message or data export, recording/transcription, deletion, webhooks, API keys, and automation as operations that need normal workplace approval, least-privilege permissions, backups where appropriate, and current Feishu documentation checks.
Forwarding messages to email can move chat content into a different security domain with weaker access controls, broader forwarding, mailbox compromise risk, and different retention behavior. The guide omits these privacy implications, which is especially dangerous in a workplace collaboration context where messages may contain confidential internal data.
The skill’s activation scope is very broad: it claims to handle essentially any Feishu-related learning, troubleshooting, or collaboration question. In an agent environment, such vague routing can cause the skill to be invoked for general requests outside its intended boundaries, increasing the chance of misrouting, over-collection of user context, or unsafe guidance on permissions, automation, or integrations without tighter scoping.
The document instructs users to obtain and copy a Webhook URL and later references API keys and Webhook integrations, but it omits any warning that these values are sensitive credentials and that external endpoints can exfiltrate chat content or trigger unauthorized messages. In a skill about Feishu operations and automation, users are likely to follow these steps directly, so the lack of security guidance materially increases the chance of credential leakage or unsafe outbound integrations.
The document gives direct steps for deleting single and multiple records, but it does not clearly warn users that deletion is destructive or that recovery is time-limited to 30 days. For markdown guidance, destructive actions that may affect user data should include explicit cautionary language.
This markdown file includes a form view that can generate share links and later explains exporting data to Excel/CSV, but it does not include any user warning about exposing records externally or handling sensitive data carefully. Because markdown files should warn about behaviors that could affect user data or privacy, the omission is a semantic safety gap.
The documentation describes starting and stopping meeting recording with automatic cloud saving, but gives no warning about participant consent, organizational policy, or legal restrictions. In a collaboration skill, this omission can normalize recording sensitive conversations without notice, creating privacy, compliance, and trust risks.
The skill presents automatic recording, transcription, summary generation, editing, sharing, search, and export as frictionless features without any privacy or confidentiality warning. Because meeting content may include personal data, internal strategy, or regulated information, this guidance could lead users to enable broad capture and distribution of sensitive material without appropriate safeguards.
This file presents all user-facing guidance exclusively in Chinese, which may constitute a language or locale policy issue when no user opt-in or regional justification is provided. There is no indication that the skill is intended only for a Chinese-speaking audience or that alternate language support is unavailable by design.
The document recommends meeting recording and automatic note generation but does not mention consent, legal notice, retention, or sensitive-data handling. In a workplace collaboration skill, this can normalize recording meetings containing personal, confidential, or regulated information without appropriate authorization or participant awareness.
The file presents all user-facing guidance in Chinese and does not indicate that users may choose another language or that the skill is intentionally limited to a Chinese-speaking audience. Under the policy criteria, forcing a specific language without user opt-in is a natural-language policy concern.
The guide instructs users how to permanently delete files and notes they are unrecoverable, but it does not add a clear caution to verify necessity, backups, or approval before destruction. In operational documentation, this can contribute to accidental irreversible data loss, especially for shared or important files.
The document explicitly teaches users to create file-sharing links with an option for public access ('任何人可见') but does not include any warning about accidental exposure of sensitive or internal files. In a file-management and collaboration skill, this omission can normalize insecure sharing practices and increase the risk of unauthorized disclosure.
This markdown file presents all instructional content in Chinese and does not indicate that users may choose another language or that the skill is intentionally limited to a Chinese-speaking audience. Under the policy, forcing a specific language without user opt-in is a natural-language policy concern.
The guide instructs users to export chat records but omits any warning that exported logs may contain sensitive personal, business, or regulated data and may be stored outside managed protections. This increases the risk of unauthorized retention, sharing, or leakage of conversation contents.
The instructions describe clearing chat history without warning that the action may be destructive, irreversible, or may hinder audits and incident review. Users may delete important evidence, operational records, or business context without understanding the consequences.
The document presents '端到端加密' as an available Feishu messaging feature without qualification, which may mislead users into believing conversations are protected at a stronger level than they actually are. In a security-sensitive collaboration tool, overstating encryption capabilities can cause users to share sensitive information under false assumptions.
The file title and all instructional content are written exclusively in Chinese, with no indication that users may choose another language or that the skill is intended only for a Chinese-speaking or region-specific audience. Under the policy, forcing a specific language without opt-in is a natural-language locale violation.
The workflow content describes automations that create records, update data, send messages, and provision enterprise accounts, but it does not warn that these actions can have real side effects on production data and user-facing systems. This is risky because readers may deploy workflows without testing, approvals, rollback plans, or permission scoping, leading to accidental data modification, spam, or unauthorized account operations.
The guidance tells users to set sharing permissions and configure authorization but does not warn about least-privilege access, oversharing, or external sharing risks. In collaboration platforms, vague permission guidance can lead to unintended disclosure of internal documents to broader audiences than intended.
该 markdown 文件适用 SQP-2 和 SQP-3，不适用 SQP-1。L417 的“确保英文输入状态”属于自然语言层面的语言/locale 约束，但没有说明仅在特定快捷键场景下临时需要，也未提供替代方案或用户选择，可能构成语言/locale policy violation。
No suspicious patterns detected.
