Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-helper skill for OpenAPI/Swagger work with no executable payload or hidden persistence, though its trigger wording is broad.
Before installing, be aware that the skill may be selected for some broad API or developer-experience requests; use it when you specifically want help with OpenAPI, Swagger, REST API documentation, schema validation, or related documentation workflows.
The trigger phrases are generic and partially malformed, making the skill easier to invoke unintentionally from ordinary user requests about OpenAPI or API docs. In an agent environment, overly broad invocation can route conversations into this skill without clear user intent, increasing the chance of inappropriate tool use or unexpected instruction application.
The trigger phrases are broad, generic, and partially templated, which increases the chance that normal user requests about APIs or documentation will invoke this skill unintentionally. In an agent environment, unintended invocation can misroute tasks, expose unrelated context to the skill, or cause the system to act on the wrong workflow, especially because this skill is designed to transform live requests into repeatable execution steps.
The skill description and use conditions are broad enough to match many common API- or software-related requests, which can cause the skill to activate outside its intended scope. Over-broad triggering is dangerous because it can route unrelated user requests into a specialized workflow, increasing the chance of irrelevant guidance, prompt hijacking exposure through unintended invocation, or unsafe downstream tool use in a larger agent system.
The example trigger phrases use very generic wording like 'Help me' and 'I need a practical workflow,' which overlaps with normal conversation and can unintentionally invoke this skill. In an agent environment, this broad matching can create misrouting and make the system easier to manipulate into selecting this skill when a more appropriate or safer skill should handle the request.
The description defines activation conditions in a vague way, combining broad topic labels with generic requests for workflows, artifacts, analysis, or implementation support. Without clear inclusion and exclusion criteria, the system may invoke this skill for loosely related software tasks, increasing the chance of misrouting and inappropriate instruction application.
The trigger keywords are overly broad and include common terms such as 'openapi', 'swagger', 'rest api', and 'developer experience', which can match many ordinary requests outside the intended scope. This can cause unintended skill activation, leading the agent to apply this skill in irrelevant contexts and potentially override better-matched instructions or workflows.
The default prompt uses a very broad activation phrase and generic task framing, which can cause the skill to be invoked in situations beyond the intended OpenAPI documentation scope. Over-broad invocation increases the chance of inappropriate routing, unintended handling of unrelated user requests, or prompt-surface expansion where attackers can more easily trigger the skill in unsafe contexts.
The trigger description and examples lack precise activation boundaries, so the skill may be selected for loosely related software requests that mention backend teams or practical workflows without actually seeking API documentation help. In an agentic system, ambiguous routing logic can create unsafe or unreliable behavior by invoking the wrong capabilities and bypassing more appropriate safeguards or specialized skills.
The trigger sentence is so generic that it can activate on ordinary user phrasing rather than clear, intentional invocation of this specific skill. Overly broad activation increases the chance of misrouting requests, causing the agent to apply the wrong workflow or expose skill behavior in contexts where it was not intended.
The file descriptions explicitly designate English and Chinese variants, but the README does not state that the user can choose their preferred language or locale. This can create an implicit language-selection policy without opt-in, which falls under language/locale policy concerns.
This README is explicitly a Chinese-language variant (zh-CN), but the file does not state that language selection is optional or user-driven. Under the language/locale policy, locale-specific behavior should either offer user opt-in or clearly document why the constraint is required.
The example trigger phrases do not establish clear boundaries and include generic wording like 'Help me' and 'I need a practical workflow', which can normalize broad matching behavior. Although examples alone are less dangerous than the main trigger logic, they still reinforce ambiguous activation patterns and may contribute to accidental invocation.
