Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overbroad activation wording, but no evidence of hidden execution, persistence, credential access, or data exfiltration.
Before installing, be aware that this skill may activate for generic web search or bug-fix requests because implicit invocation is enabled and the triggers are broad. It appears safe from an execution and data-access standpoint, but users who want precise routing should narrow the triggers or invoke it explicitly by name.
The skill description is extremely broad and includes generic work terms plus multiple adjacent concepts, which can cause the skill to be selected in situations only loosely related to Tavily/web-search workflows. Unintended invocation can route user requests into the wrong workflow, causing confused-deputy behavior, irrelevant actions, or unsafe assumptions in downstream agent behavior.
The trigger keyword list contains highly generic tokens such as 'web', 'search', 'api', 'user', and 'bug fix', which are common across many unrelated requests. In an agent routing system, these vague triggers can cause frequent over-matching, leading this skill to intercept unrelated tasks and increasing the chance of incorrect guidance or execution paths.
The trigger keywords are overly broad and include generic terms such as 'web', 'search', 'api', and 'user', which can match many unrelated requests. This can cause unintended activation of the skill, expanding its operational scope beyond user intent and creating prompt-routing confusion that may expose users to irrelevant or lower-safety workflows.
The trigger sentence begins with a very broad everyday phrase ('Help me ...') that can match many unrelated user requests and cause unintended skill activation. In an agent environment, over-broad routing can expose users to irrelevant web-search workflows, increase prompt-surface area, and interfere with correct tool or skill selection.
The 'I need a practical workflow for ...' pattern is also overly general and lacks strong scope restriction, making accidental activation likely for many generic productivity requests. Because this skill is framed broadly around workflows, bugs, hardening, and adjacent implementations, an unrestricted trigger can capture requests outside its intended domain and degrade safety and reliability of agent behavior.
The trigger sentences are broad, repetitive, and phrased in a way that can match loosely related requests, increasing the chance the skill is invoked when the user did not clearly intend it. In an agent ecosystem, unintended invocation can route user tasks into the wrong workflow, causing incorrect actions, wasted tool usage, or unnecessary exposure of user context to web-search-oriented processing.
The trigger phrases are broad, generic, and overlap with ordinary user requests such as asking for help, workflows, or bug fixes. This can cause the skill to activate outside its intended scope, leading to prompt/skill routing confusion and unintended execution of search-oriented workflow guidance in unrelated contexts.
The example trigger phrases are vague and effectively mirror broad descriptive text rather than demonstrating precise activation boundaries. This makes it harder for maintainers and routing logic to distinguish intended use from incidental phrase overlap, increasing accidental invocation risk.
The description says to use the skill when users ask for broad categories like work-productivity, web, or search, or when they need practical workflow support, which is vague and expansive. In a skill-routing context, ambiguous invocation conditions increase the chance of accidental selection and inappropriate application of this skill to unrelated tasks.
The default prompt uses a very broad activation phrase tied to common words like "help me" and a generic productivity/search workflow description, which can cause the skill to trigger on ordinary user requests unrelated to explicit intent to use this skill. Because implicit invocation is enabled, this increases the chance of accidental routing, prompt interference, and untrusted skill behavior being injected into normal conversations.
The natural-language content includes Chinese text inside an otherwise English README, but the file does not state whether the skill is intended to operate bilingually or let the user choose language. Under the language/locale policy, forcing or assuming a language without clear opt-in can be a policy concern.
This user-facing README presents substantial English content inside a Chinese-language document, but it does not explain whether the skill responds in Chinese, English, or adapts to user preference. That can create a language-policy issue if the skill effectively imposes a language or locale without clear opt-in or choice.
