Security audit
Security checks for vulnerabilities and agentic risk
This skill is a Chinese-language project-planning template skill that does not request code execution, credentials, persistence, or privileged access.
Install this if you want a Chinese-language project planning and task decomposition assistant. Be aware that generic planning requests may trigger it, and rendering the README may load an external image from qomob.ai.
The trigger description is overly broad and includes common phrases like '帮我拆解任务' and '做个项目计划', making accidental invocation likely during ordinary conversation. In agent ecosystems, overbroad routing can cause this skill to intercept unrelated requests, override more appropriate skills, and steer user interactions into an unintended workflow, which is a prompt-scope and control-boundary issue.
The trigger phrases are broad enough to match common planning requests such as 'help me break down a task' or 'make a project plan,' which risks the skill activating in contexts the user did not specifically intend. Over-broad invocation can override other more appropriate skills, capture unrelated user inputs, and increase exposure to prompt-scope confusion.
The skill metadata and content are Chinese-only and do not offer language negotiation or justify a locale restriction, which can force behavior misaligned with the user's language and context. In a multi-skill or multilingual environment, this can lead to user confusion, incorrect requirement capture, and unsafe downstream planning decisions because constraints and deliverables may be misunderstood.
This markdown file uses a single language throughout and does not indicate that users may choose another language or locale. Under the natural-language policy rule, forcing a specific language without user opt-in is a policy concern unless the locale restriction is documented and justified.
Suspicious Unicode normalization or mixed-script content
The natural-language instructions and examples are entirely in Chinese, and the document does not indicate that users may choose another language or locale. Under the stated policy, forcing a specific language without user opt-in can be a language/locale policy violation.
The README embeds an externally hosted image, which causes clients rendering the markdown to contact a third-party domain. That can leak viewer IP address, user agent, referrer/workspace context, and creates a passive tracking surface unrelated to the skill’s task-planning function.
The changelog and v4.1 version record say the skill removed the '输入验证门/InputGate' in v4.1, but the embedded workflow diagram still includes an '[Input Validation Gate]'. This is an active contradiction in the skill's own documentation about how the skill operates, even though it does not introduce a new capability.
This markdown file contains user-facing instructional content exclusively in Chinese, and nowhere indicates that users may choose another language or that the template is intentionally limited to a Chinese-speaking audience. Under the policy rule for language or locale constraints, forcing a specific language without opt-in is a natural-language policy concern.
No suspicious patterns detected.
