Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with no code execution, credentials, persistence, or data exfiltration, though its automatic activation triggers are too broad.
Install only if you want a helper for skill-vetting workflow design. Be aware that its implicit invocation keywords are broad, so you may want to disable implicit invocation or invoke it explicitly by name to avoid unrelated requests being handled by this skill.
The skill description includes very broad trigger terms such as 'security' and 'first', which are common in many unrelated conversations. This can cause unintended invocation of the skill, leading the agent to activate workflows outside the user's intent and potentially inject security-vetting behavior into unrelated tasks.
The keyword list contains multiple vague everyday words, including 'security', 'first', 'before', and 'github', without any scoping qualifiers. In an agent system, such generic triggers can cause accidental routing to this skill during many unrelated requests, creating misfires, confusion, and possible interference with safer or more appropriate workflows.
The trigger list includes highly generic terms such as "security", "first", "before", "github", and "bug fix", which can cause this skill to activate in many unrelated conversations. Over-broad activation increases the chance of unintended skill invocation, context hijacking, and irrelevant guidance being injected into workflows where a more appropriate skill should handle the request.
The trigger sentences and keywords are broad enough to match common requests such as 'security', 'github', 'before', or 'bug fix', which can cause the skill to activate outside its intended scope. In an agent ecosystem, this creates routing ambiguity and can divert unrelated user tasks into this workflow, increasing the chance of incorrect handling, unintended disclosure of context, or policy bypass through over-selection of this skill.
The trigger phrases are broad, generic help-seeking language that can cause the skill to activate in situations unrelated to security vetting or this specific workflow. In a security-oriented skill, overbroad activation increases the chance of unsolicited interception of user intent, misrouting sensitive tasks, or unexpectedly applying this skill where a narrower, more appropriate skill should be used.
The file name and headings indicate a zh-CN version, yet multiple core user-facing sections, including requirements, evidence descriptions, workflow description, and trigger phrases, are written in English. This creates an implicit language constraint mismatch for users expecting Chinese content and does not provide an opt-in or language selection.
The trigger phrases are broad and include generic terms like security, first, before, installing, and github, which can cause the skill to activate in many unrelated conversations. In an agent ecosystem, unintended invocation can route user requests into the wrong workflow, causing confusing behavior, mis-scoped actions, or accidental processing of sensitive repository or security-related contexts.
The example trigger sentences reinforce ambiguous activation by using incomplete, low-specificity phrasing and repeating broad concepts rather than clear user intents. This trains or signals a loose activation boundary, making it more likely that the skill will be invoked for partial phrase matches or unrelated requests.
The activation description says the skill should be used for broad categories like work-productivity, security, or practical workflow support, without sufficiently constraining scope. This ambiguity can lead to accidental invocation across many unrelated requests, making routing less reliable and potentially exposing users to irrelevant or conflicting operational guidance.
The default prompt contains broad, generic trigger language such as 'help me' and references common topics like productivity, security, workflow, and implementation support. Because implicit invocation is enabled, this can cause the skill to activate in response to ordinary user requests that were not clearly intended for this specific skill, increasing the chance of unintended prompt injection exposure, context leakage into the skill, or misrouting of user tasks.
The file lists English and Chinese variants for instructions and guides, but it does not state how the user's preferred language is selected or that language choice is optional. This can create a locale-policy concern if the skill defaults users into a language without explicit opt-in.
