Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with overly broad activation wording, but it does not add code execution, credentials, persistence, or hidden behavior.
Install only if you want a broad ontology/productivity workflow helper. Be aware it may activate too easily because implicit invocation is enabled and the trigger wording is generic; prefer tighter triggers before using it in environments where skill routing accuracy matters.
The skill description and 'Use when' text are broad enough to match many ordinary productivity, knowledge, or implementation requests that are not specific to this skill's actual scope. In an agentic system, this can cause incorrect auto-selection, route user prompts into the wrong skill, and expand the influence of this prompt over unrelated tasks.
The trigger keywords include highly generic terms such as 'knowledge', 'graph', 'structured', 'memory', 'creating', and 'bug fix', which can match a very large fraction of unrelated prompts. This makes overbroad invocation likely and increases the chance that the skill intercepts requests outside its intended domain, degrading reliability and potentially applying inappropriate workflow instructions to sensitive tasks.
The skill enables implicit invocation without any narrowly defined trigger conditions, which means ordinary user requests about productivity, workflows, knowledge graphs, or checklists could cause the agent to invoke this skill unexpectedly. In practice, this broad auto-routing increases the attack surface for prompt injection, unintended tool activation, and scope creep because the skill can be selected in many ambiguous contexts.
The trigger sentence is broad enough that ordinary user requests could accidentally activate this skill outside its intended scope. In an agent routing system, overbroad activation can cause misselection, unexpected behavior, and unsafe delegation if the skill handles tasks it was not meant to process.
The trigger phrases are broad enough to match common productivity and help-seeking requests, which can cause the skill to activate outside its intended scope. In an agent ecosystem, overbroad routing can misapply instructions or workflows to unrelated tasks, increasing the chance of unsafe automation, user confusion, or unintended data handling.
This zh-CN README presents substantial user-facing content, including the main requirement description, workflow explanation, and trigger phrases, in English rather than the document's stated Chinese locale. That effectively imposes a language mismatch on users without opt-in or an explicit bilingual policy.
The activation section uses generic phrases like "Help me" and "I need a practical workflow" before a long copied description, which creates ambiguity about what should actually trigger the skill. The file does not provide clear trigger constraints, narrow context, or negative examples to prevent accidental invocation from ordinary conversation.
The example trigger sentences are malformed, repetitive, and do not establish clear activation boundaries, so they fail to constrain routing behavior and may reinforce accidental matching on copied generic language. Poor trigger examples are dangerous in skills because they act as informal routing guidance for maintainers and automated selectors, increasing misfires and unintended prompt capture.
The skill description defines applicability very broadly across work-productivity, ontology, typed, knowledge, graph, and any request for practical workflows, artifacts, checklists, analysis, or implementation support. This unclear boundary can make the skill match a large class of normal requests, causing misrouting and unintended invocation in contexts where the skill is not the best or safest fit.
The trigger keyword list includes broad generic terms such as 'structured', 'memory', 'creating', and 'bug fix', which can overlap with many unrelated user requests and cause the skill to activate outside its intended scope. In an agent environment, over-triggering can route users into irrelevant workflows, increase prompt injection exposure surface, and interfere with safer or more appropriate skills.
The default prompt uses very broad everyday language such as work-productivity, workflow, checklist, analysis, and implementation support, which overlaps heavily with common user intent. This makes accidental or overly broad invocation more likely, especially when combined with implicit invocation, and can cause the agent to route requests to this skill outside its intended scope.
Although this trigger includes more wording, it still matches common everyday phrasing like 'I need a practical workflow,' which is too generic for safe skill routing. This increases the chance of accidental invocation, causing the agent to prioritize this skill for unrelated tasks and potentially bypass more appropriate or safer skills.
该技能文件为 SKILL.zh-CN.md，内容整体以简体中文提供，但未在正文中说明这是可选语言版本，也未提示用户可选择其他语言。按规则，若技能对语言/区域有强制性而缺少用户选择或正当说明，可能构成自然语言策略问题。
No suspicious patterns detected.
