Security audit
Security checks for vulnerabilities and agentic risk
This is a benign documentation-only unit-test coverage helper, with only a routing-quality concern that broad implicit triggers could make it activate too often.
Before installing, be aware that the skill may be invoked for general testing or quality requests because its trigger wording is broad. It is otherwise a straightforward local unit-test coverage workflow with no executable payload or persistence.
The invocation guidance lacks precise boundaries for when the skill should or should not run, which can lead to accidental or overly frequent activation. In an agent environment, ambiguous routing is a security-relevant weakness because it can cause unintended instruction precedence, context mixing, or execution of the wrong workflow for a user's request.
The trigger sentence is so generic that it can match ordinary user requests and cause the skill to activate outside a clearly intended scope. Over-broad activation increases the chance of prompt-routing mistakes, unnecessary exposure of user context to the skill, and interference with more appropriate skills or baseline handling.
The trigger phrases are broad and can activate the skill for generic requests about software, testing, or workflows without sufficient constraints. In an agent environment, overbroad activation can cause the wrong skill to take control, produce irrelevant actions, or expose repository context unnecessarily, which increases the chance of unsafe or unintended behavior.
The skill description is broad enough to match many ordinary software requests, which can cause the agent to invoke this skill outside its intended scope. Over-broad routing increases the chance of prompt/context capture, where a generic skill inserts its workflow and assumptions into unrelated coding tasks, reducing safety and reliability.
The keyword list uses generic terms like 'testing', 'regression', and 'quality' without qualifiers, making accidental triggering likely across a wide range of software conversations. This can cause inappropriate skill activation and unwanted instruction injection into sessions that were not actually asking for test-coverage help.
The skill description and usage conditions are broad enough that an orchestrator could invoke this skill for generic software or testing requests outside its intended scope. Over-broad triggering can cause misrouting, unexpected tool behavior, or disclosure of repository context to an irrelevant skill, which is a real security and reliability concern in agent systems.
The keyword list includes generic terms like testing, regression, and quality without contextual qualifiers, which increases the chance of accidental activation. In an agent environment, broad keywords can route unrelated prompts into this skill, leading to confused delegation, unnecessary code-oriented actions, or exposure of project details to a workflow the user did not request.
The default prompt uses a very broad natural-language activation phrase tied to common testing and software help requests, which increases the chance the skill is invoked in situations the user did not explicitly intend. In an agent environment, overbroad invocation can cause prompt hijacking of normal workflows, unnecessary context exposure to the skill, or user confusion about which capability is acting.
Implicit invocation is enabled without documented trigger constraints, exclusions, or scope boundaries, so the platform may auto-route ordinary development conversations into this skill. That can lead to unintended skill activation, leakage of repository or conversation context into the skill, and interference with other more appropriate tools or policies.
The trigger sentence is broad enough to match ordinary user phrasing, which can cause the skill to activate in contexts beyond its intended scope. In an agent system, overly broad routing can lead to inappropriate context injection, misapplied workflows, or unintended handling of unrelated requests, increasing the chance of downstream security or reliability issues.
The example triggers are phrased so generally that they normalize activation from vague, everyday requests rather than specific coverage-related intents. This reinforces over-triggering behavior and can train routing systems or authors to invoke the skill in ambiguous contexts.
L46-L50 给出了触发示例，但示例基本是模板化的“Help me/I need...”句式，未展示与其他测试相关请求的边界，也没有说明哪些类似表达不应触发该技能。缺少负例会使调用范围显得含糊。
No suspicious patterns detected.
