Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording, but no hidden execution, persistence, credential handling, or data exfiltration behavior.
Before installing, be aware that this skill may activate too often because its keywords are broad and implicit invocation is enabled. It is otherwise a low-risk workflow/documentation helper; consider narrowing triggers if you maintain it.
The trigger phrases are so generic that they can cause the skill to activate for unrelated requests containing common terms like 'security', 'first', 'github', or 'bug fix'. In an agent ecosystem, overly broad activation increases the chance of unintended routing, inappropriate application of this workflow, and user confusion that can lead to unsafe or incorrect task handling.
The manifest description includes very broad activation terms such as "security" and "first," which can cause the skill to match many ordinary conversations unrelated to this specialized workflow. Over-broad activation increases the chance the skill is invoked unexpectedly, creating prompt-routing confusion, accidental disclosure of context to an irrelevant skill, and reduced user control over which skill is active.
The trigger keyword list contains ambiguous everyday words including "security," "first," "before," and "github," none of which are constrained to the skill's intended domain. This makes accidental activation highly likely across normal user requests, potentially hijacking unrelated tasks and causing the agent to apply the wrong workflow or expose user context to an unnecessary skill.
The trigger list includes overly broad everyday terms such as “security”, “first”, “before”, and “github”, which can cause the skill to activate in many unrelated contexts. In an agent environment, this can lead to incorrect routing, unintended invocation of this workflow, and interference with more appropriate skills or user intent handling.
The trigger sentences and keywords are broad enough to match common user requests such as 'security', 'github', 'first', or generic asks for a 'practical workflow'. This can cause the skill to activate outside its intended niche, creating prompt-routing confusion and potentially applying security-vetting behavior to unrelated requests, which weakens least-privilege skill selection and can distort downstream agent behavior.
The trigger phrases are broad enough to match common terms like 'security', 'github', 'first', and 'before', which can cause the skill to activate in contexts far beyond its intended use. Over-broad activation increases the chance of unintended invocation, context hijacking, and user confusion, especially for a skill positioned around vetting and workflow guidance where precision matters.
The example trigger sentences are malformed and do not establish a clear, bounded activation condition, which undermines reliable routing and makes it harder to distinguish intended from incidental matches. In combination with already broad keywords, these vague examples reinforce over-triggering and inconsistent behavior rather than safely constraining use.
The description says to use the skill when a user asks for broad topics like work-productivity, security, or practical workflow support, without clear boundaries. This ambiguity increases the chance of accidental selection for unrelated requests, which can degrade reliability and create unsafe automation behavior through misclassification rather than direct code execution.
The default prompt uses broad, generic phrasing such as 'help me' and references common tasks like fixing bugs, hardening setup, reliability, checklists, analysis, and implementation support. This overlaps with ordinary user requests, increasing the chance that the platform will route unrelated conversations into this skill, especially when combined with implicit invocation.
The skill enables allow_implicit_invocation without any narrowly defined activation constraints, which means the system may invoke this skill on broad user requests without clear user intent. Because the skill is framed around security, vetting, and workflow help, accidental invocation could expose users to unintended autonomous behavior, prompt interference, or skill overreach in unrelated conversations.
The README states that SKILL.md and README.md are English and the corresponding .zh-CN files are Chinese, but it does not describe how the user's preferred language is selected or whether language choice is optional. This can create a locale-policy concern if the skill defaults users into a language without explicit opt-in.
