Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only product validation helper with overly broad triggers but no evidence of code execution, credential access, persistence, or destructive behavior.
Install this if you want structured product validation and MVP planning help. Be aware that its broad implicit triggers may activate during loosely related startup, SaaS, prototype, or validation discussions; explicit invocation or tighter triggers would make routing clearer.
The trigger sentence is extremely broad and can match common product-help requests, causing the skill to activate in situations far beyond its intended scope. Over-broad activation can route unrelated conversations into this skill, leading to prompt hijacking of task selection, degraded safety-policy routing, or inappropriate disclosure/processing under the wrong workflow.
The trigger phrases are generic and map to broad business/product terms like 'validation', 'prototype', and 'saas', which increases the chance the skill is invoked for loosely related requests rather than clear user intent. In an agent ecosystem, over-broad routing can cause unintended activation, leading to irrelevant guidance, confused user experience, and possible exposure to downstream risky instructions contained in the skill files.
The trigger phrases are broad enough to match common product, business, and workflow requests, which can cause the skill to activate in situations beyond its intended scope. Over-broad activation increases the chance of prompt-routing mistakes, unintended instruction injection from loosely related requests, and user confusion about why this skill was selected.
The skill description is broad enough to match common product, business, and implementation requests that may not actually need this skill. Overbroad matching can cause unintended invocation, leading the agent to apply the wrong workflow, make incorrect assumptions, or overshadow more appropriate skills in multi-skill environments.
The trigger keywords and example phrases are highly generic terms like 'validation,' 'prototype,' 'saas,' and 'startup,' which are likely to appear in many unrelated requests. This increases the chance of accidental routing and misuse of the skill, especially because the examples encourage invocation from broad natural-language matches rather than tightly scoped intents.
The top-level invocation guidance says to use the skill for broad categories like business operations, product ideas, prototypes, and SaaS support, without strong scoping constraints. This can lead to accidental invocation across a wide range of benign conversations, causing unintended instruction precedence and reducing user control over which specialized workflow is applied.
The file presents the skill instructions and operating guidance in Chinese and is specifically named as a zh-CN variant, but it does not state that language choice is optional or explain a justified region-specific requirement. Under the policy, forcing a specific language without user opt-in is a natural-language policy concern.
The trigger keywords and example invocations are broad enough to match many ordinary product or startup discussions, which can cause the skill to activate outside its intended scope. Over-broad activation increases the chance of prompt/context hijacking at the routing layer, where this skill could inappropriately shape responses for unrelated requests.
The default prompt uses a very broad invocation phrase and expansive problem statement, which can cause the agent to activate in situations beyond narrowly intended product-validation use cases. Because implicit invocation is enabled, this increases the chance of over-triggering, unintended data exposure to the skill, or the agent influencing workflows where it was not explicitly requested.
This trigger relies on an ambiguous, general request for a 'practical workflow,' which overlaps with many benign tasks outside product validation. Ambiguous activation criteria increase the chance of unintended invocation, making orchestration less predictable and potentially bypassing more appropriate specialized skills or review paths.
This Chinese README includes key user-facing operational text and trigger phrases in English, which can impose an implicit language expectation on users without opt-in or explanation. The file does not state that English input is required or offer an explicit language choice.
No suspicious patterns detected.
