Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with no executable payload, persistence, credential handling, or hidden data movement, though its broad trigger wording may cause it to activate too often.
Before installing, consider narrowing or disabling implicit activation if your agent selects skills automatically, because common terms like security, github, or bug fix could invoke this helper for unrelated work. There is no evidence here of hidden execution, exfiltration, credential use, or persistence.
The description says to use the skill when a user asks for broad categories like "work-productivity," "security," or any practical workflow or checklist support, which makes invocation boundaries ambiguous. This increases the chance that the skill is selected for unrelated tasks, leading to incorrect handling, missed domain-specific safeguards, or routing away from a more appropriate skill.
The trigger list includes very broad, common terms such as "security," "first," "before," and "github," which can match many unrelated user requests and cause unintended activation of this skill. In an agent system, over-broad routing can misapply the wrong workflow, confuse users, and in security-sensitive contexts may divert tasks to incomplete or misleading guidance.
The trigger keywords are overly broad and include common terms such as "security", "first", "before", "github", and "bug fix", which can overlap with many unrelated user requests. This creates a real risk of unintended activation, causing the agent to inject this workflow into contexts where it is not appropriate and potentially override more relevant or safer skills.
The trigger phrases are broad, natural-language fragments such as 'security', 'first', 'before', and generic help-me formulations, which can cause the skill to activate in unrelated conversations. In an agent ecosystem, unintended invocation can route user requests into the wrong workflow, creating security-review confusion, noisy automation, or inappropriate handling of tasks that were not meant for this skill.
The trigger phrases are broad enough to match ordinary requests about security, GitHub, bug fixes, or general workflow help, which can cause the skill to activate outside its intended scope. In an agent ecosystem, overbroad auto-invocation can steer unrelated conversations into this skill's workflow, creating prompt-routing errors and increasing the chance of unintended actions or misleading security guidance.
The trigger phrases are broad and include generic security and workflow terms like "security", "first", "before", and "github", which can cause the skill to activate in many unrelated contexts. In an agent environment, ambiguous activation increases the chance of unintended routing, causing the model to apply this skill when the user did not request it and potentially influencing security-sensitive guidance or analysis unexpectedly.
该技能文件整体为 zh-CN 中文版本，内容中未见向用户提供语言选择、回退机制或明确说明仅适用于特定中文用户群体。按规则，若技能强制特定语言而无用户选择或正当限定，可能构成语言/locale 政策问题。
The description says to use the skill when users mention broad concepts like work-productivity, security, or need a checklist, analysis, or implementation support, but it does not clearly define boundaries for activation. This ambiguity can lead to misrouting and accidental invocation on general-purpose requests, which is a genuine safety and reliability issue in an agent skill selection system.
The default prompt is broadly phrased and can activate on generic terms like 'work-productivity,' 'security,' 'analysis,' or 'implementation support,' which are not narrowly scoped to a specific vetted task. Combined with implicit invocation, this can cause the skill to be triggered in contexts the user did not clearly intend, increasing the chance of prompt injection exposure, accidental delegation, or unintended handling of sensitive tasks.
The README advertises multiple language-specific skill files but does not indicate any user choice, consent, or selection behavior. This can lead to the wrong language variant being applied implicitly, which may confuse users and, in a security-oriented workflow, increase the risk that warnings, constraints, or remediation steps are misunderstood.
