Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording but no code execution, persistence, credential access, or hidden data handling.
Installers should expect a low-risk workflow helper, but should consider narrowing its trigger keywords or disabling implicit invocation if they do not want it selected for general security, GitHub, or bug-fix requests.
The manifest description uses very broad activation language such as "Use when a user asks for work-productivity, skill-vetter, vetter, security, first," which creates ambiguous routing boundaries. In agent ecosystems, this increases the chance that the skill is selected for unrelated conversations, exposing users to irrelevant instructions and interfering with safer or more appropriate skills.
The trigger keywords are excessively broad, including generic terms like "security," "first," "before," and "github," which are likely to match many unrelated user requests. This can cause accidental activation of the skill outside its intended scope, leading to prompt/context hijacking of normal workflows and unexpected behavior in agent routing.
The trigger keywords include very broad, common terms such as "security", "first", "before", and "github", which can cause the skill to activate in many unrelated conversations. This increases the chance of accidental invocation, context hijacking, and inappropriate influence over tasks the user did not intend this skill to handle.
The trigger sentences and keywords are broad enough to match many ordinary requests involving security, GitHub, bug fixes, or general workflow help, which can cause the skill to activate outside its intended niche. In an agent ecosystem, over-broad routing can misdirect users into an inappropriate workflow, increasing the chance of irrelevant guidance, prompt-scope confusion, or unintended handling of security-related tasks.
The trigger phrases are overly generic and can match ordinary user requests about security, GitHub, bug fixing, or practical workflows, causing the skill to activate outside its intended scope. In an agent ecosystem, this can lead to unintended routing, misleading assistance, or inappropriate application of this skill in contexts where a more specialized or safer skill should handle the request.
The trigger phrases are broad enough to match common requests such as security help, GitHub help, bug fixing, or generic workflow assistance, which can cause the skill to activate outside its intended scope. In an agent ecosystem, this increases the chance of unintended routing, confusing users, or inserting this skill into sensitive security-related workflows where its generic behavior may override a more appropriate specialized skill.
The example trigger sentences are vague, unnatural, and non-constraining, so they do not meaningfully teach a router or maintainer when the skill should be used. Poor trigger examples reinforce overbroad matching and make accidental invocation more likely, especially when combined with the already generic keyword set.
The skill description defines applicability very broadly, covering many user types and a wide range of tasks like bug fixing, safety hardening, reliability improvement, and adjacent workflow creation. Without clearer boundaries, the orchestrator may select this skill for loosely related requests, leading to overreach, unintended handling of sensitive security topics, or interference with more appropriate skills.
The default prompt invokes the skill with broad, common-language terms such as "help me" and a wide functional scope, which can cause the agent to select this skill in situations the user did not clearly intend. Because implicit invocation is enabled, this increases the chance of over-triggering and unintended handling of security- or workflow-related requests, which can misroute sensitive tasks or bypass more appropriate skills.
This Chinese README contains substantial English content in core descriptive sections, which can impose a language preference on users of the localized file. There is no indication that users may choose their preferred language or that the mixed-language presentation is intentional and justified.
