Security audit
Security checks for vulnerabilities and agentic risk
This skill is disclosed as Douyin automation, but it enables comments, private messages, and likes with weak recipient-safety controls and limited scoping.
Review this carefully before installing. Use it only for accounts and recipients where you have permission to interact, keep sending actions manual or confirmation-gated, add strict rate limits, and verify Douyin platform rules. Be aware that automated comments, DMs, and likes can affect other users and may risk account enforcement.
The skill presents itself as an operational Douyin automation package, yet the content shown is primarily instructional text and browser action examples, not a complete executable skill. In a security context, this kind of overclaiming reduces transparency and may conceal recipient-facing automation or additional code paths not obvious to reviewers.
The skill presents itself as an operational Douyin automation package, yet the content shown is primarily instructional text and browser action examples, not a complete executable skill. In a security context, this kind of overclaiming reduces transparency and may conceal recipient-facing automation or additional code paths not obvious to reviewers.
The README is written entirely in Chinese and prescribes fixed Chinese interaction style guidance such as '不自我介绍，纯正常沟通' without indicating that users may choose another language or that the skill is intentionally limited to a China-specific locale. This can violate language/locale policy when no opt-in or justification is provided.
The README explicitly promotes automated commenting and private messaging while only advising operators to add random delays and be cautious, but it does not meaningfully warn about spam, privacy, consent, or account-enforcement risks. In this context, the omission is dangerous because the skill is designed for outreach automation on a consumer platform, which can facilitate unsolicited contact, user harassment, or account suspension at scale.
Without declared permissions the skill's intent is opaque and cannot be validated.
The description and examples are entirely prescriptive in Chinese and direct the interaction style for outputs, but the document does not state that the skill is intentionally limited to a Chinese-language Douyin context or allow user opt-in for language/locale behavior. Under the policy, forcing a specific language or locale without choice or justification is a natural-language policy concern.
The skill explicitly enables automated commenting and private messaging to third parties while normalizing 'pure normal interaction' and 'verified' status, but it does not warn about unsolicited outreach, spam risk, account sanctions, or recipient harm. In this context, the functionality is inherently recipient-facing and can easily be used for mass outreach or deceptive engagement.
The module description and all built-in reply phrases are written exclusively in Chinese, and the skill is described as a generic auto-replier rather than a clearly region-specific compliance tool. This creates a natural-language locale policy issue because it forces one language without explicit user opt-in or documented justification.
The reset command irreversibly overwrites the reply-history file without confirmation, backup, or warning. In an automation context, accidental or scripted invocation can erase state used to prevent duplicate replies, causing operational abuse, audit loss, or repeated messaging behavior.
This Python file includes a DM automation workflow that navigates to Douyin, enters text, and clicks send. While it notes platform risk control, it does not disclose to the user that the action will transmit user-provided content to an external service, which is a safety-relevant networked action for a code file.
The manifest focuses on Douyin interaction actions such as replying, sending DMs, liking, and searching topics. This file additionally stores reply history to a local JSON file and later exposes that state for operational control, which is behavior beyond the user-facing automation actions described in the manifest.
The manifest presents the skill as an automation package for Douyin interactions, but this module also provides administrative capabilities to inspect reply counts and reset stored history. Those maintenance operations are not mentioned in the manifest description and go beyond the listed interaction functions.
All natural-language instructions and workflow descriptions in the file are presented only in Chinese. Under the policy, forcing a specific language without user opt-in can be a locale/language policy violation unless the constraint is explicitly justified as region-specific.
No suspicious patterns detected.
