Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with no executable code, persistence, credential handling, or hidden data movement, though its activation keywords are too broad.
Before installing, consider narrowing or disabling implicit activation so this helper only runs for explicit skill-vetting workflow requests. No evidence of malicious behavior, exfiltration, or local system modification was found.
The skill description includes very broad activation terms such as "security" and "first," which are common in unrelated user requests and can cause unintended invocation. In an agent environment, accidental routing to the wrong skill can distort task handling, bypass more appropriate safeguards, or create confusing security-adjacent behavior at scale.
The keyword list contains vague everyday words including "security," "first," "before," and "github," which are likely to match a wide range of normal conversations unrelated to this skill. This makes unintended invocation highly probable, and because the skill operates in a security/vetting context, misrouting could interfere with other workflows or cause users to receive irrelevant or overly authoritative guidance.
The trigger sentences and keywords are broad enough to match many ordinary requests such as generic security, GitHub, bug-fix, or workflow queries. That can cause the skill to activate outside its intended scope, leading users or downstream agents into applying this skill in unrelated contexts and increasing the chance of incorrect guidance, prompt hijacking surface, or policy bypass through misrouting.
The trigger sentence is broad enough to match ordinary user requests about security, GitHub, bug fixes, or practical workflows, which can cause the skill to activate outside its intended scope. In an agent ecosystem, overbroad invocation can route unrelated conversations into this skill unexpectedly, increasing the chance of incorrect actions, prompt interference, or accidental use of security-themed guidance where it was not requested.
The trigger phrases are broad enough to activate on generic terms like 'security', 'github', or 'bug fix', which can cause the skill to run in unintended contexts. In an agent ecosystem, overbroad activation can misroute user requests, override a more appropriate skill, or insert this workflow into sensitive tasks where its assumptions do not fit.
The description says to use the skill whenever a user asks for broad categories like work-productivity, security, or practical workflow support for the requirement, but it does not clearly constrain scope. This ambiguity increases the chance of accidental invocation across unrelated tasks, which can misroute agent behavior and weaken predictable skill selection.
The trigger list includes very broad everyday terms such as "security", "first", "before", and "github", which can cause this skill to activate in many unrelated conversations. Over-broad activation can hijack routing, causing users to receive irrelevant workflow guidance or bypass a more appropriate specialized skill, which is a real security/safety concern in agent selection logic.
The default prompt contains very broad trigger terms such as 'help me' plus generic productivity and security-related language, while implicit invocation is enabled. This creates a realistic risk of accidental routing during ordinary conversation, causing the skill to activate without clear user intent and potentially inject workflow behavior or analysis into unrelated tasks.
The file is labeled as a Simplified Chinese README, but substantial instructional content and trigger phrases are written in English. This can amount to a language/locale policy issue because the documentation implicitly forces mixed-language usage without user opt-in or justification.
该文件为 SKILL.zh-CN.md，全文以简体中文（zh-CN）呈现，但未说明这是可选语言版本、未提供用户语言选择，也未解释该语言限制的适用范围。根据规则，若技能强制特定语言而无用户选择或合理、明确的区域性说明，可构成自然语言政策问题。
