Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only workflow helper with broad auto-activation wording, but no evidence of malicious code, hidden actions, credential use, persistence, or data exfiltration.
This skill appears safe to install from a security perspective, but users should be aware it may activate too broadly for general workflow, task, or bug-fix requests. Prefer explicit invocation when you want this specific proactive-agent workflow helper.
The manifest description is very broad, covering generic requests for "practical workflow, artifact, checklist, analysis, or implementation support" tied to common productivity terms. An underspecified invocation scope increases the chance this skill is selected for unrelated tasks, which can misroute user requests and let this skill's instructions override more appropriate, narrower skills.
The trigger list includes generic terms like "proactive," "transform," "task," "needs," and "bug fix," which are common in unrelated user requests. This can cause unintended activation of the skill, leading to prompt/scope hijacking at the routing layer where the wrong skill influences responses in contexts the user did not intend.
The trigger list includes very generic terms such as "proactive", "task", "needs", and "bug fix", which are common in ordinary conversations and likely to cause accidental activation. This can route unrelated user requests into this skill unexpectedly, increasing prompt-surface exposure and causing the agent to apply workflow behavior outside its intended scope.
The default prompt and description use very broad activation language tied to generic terms like work-productivity, proactive-agent, workflow, checklist, analysis, and implementation support. This can cause the skill to be selected in situations far outside its intended scope, increasing the chance of unintended prompt injection exposure, misleading assistance, or unsafe autonomous workflow execution in unrelated contexts.
Enabling implicit invocation without a tightly constrained activation scope allows the agent platform to auto-select this skill based on vague similarity matches rather than clear user intent. In combination with the broad productivity-oriented description, this raises the likelihood of unrequested activation, cross-context interference, and execution of the skill in sensitive workflows where its guidance may be inappropriate or unsafe.
The trigger sentence is written as a broad, everyday-style request that could match many unrelated user prompts, causing the skill to activate outside its intended scope. In an agent workflow context, unintended invocation can lead to wrong-task execution, confusing outputs, or accidental application of proactive workflows where the user did not request them.
The trigger phrases are broad, generic, and partially mirror common user intents such as asking for a practical workflow or help with tasks. This can cause unintended invocation of the skill in unrelated contexts, increasing the chance that users are routed into the wrong workflow and that downstream instructions are applied when the user did not explicitly request this skill.
The trigger phrases are broad and generic enough to match common user requests like asking for practical workflows, task help, or bug fixes. This can cause the skill to activate outside its intended scope, leading to inappropriate interception of unrelated requests and increasing the chance that users are routed into an unneeded or misleading workflow.
The description says to use the skill when a user asks for broad categories like work-productivity, proactive-agent, proactive, transform, or task, without defining clear inclusion and exclusion criteria. That ambiguity can make orchestration systems invoke the skill in situations only loosely related to its purpose, reducing predictability and increasing the chance of inappropriate handling of user requests.
This trigger is underspecified and lacks clear activation boundaries, which increases the risk of false-positive routing. Because the skill is framed as practical workflow support, accidental selection could steer conversations, analyses, or implementation steps inappropriately and reduce safety by bypassing more suitable skills or normal user clarification.
The file descriptions state that SKILL.md contains English instructions and SKILL.zh-CN.md contains Chinese instructions, but the README does not explain whether users can choose their preferred language at runtime. This can create a locale-policy concern if the skill defaults behavior or outputs based on fixed language-specific instructions without user choice.
The evidence section includes multiple Chinese-language titles and text alongside English content, but the document does not indicate whether multilingual output is optional or explain any locale-specific requirement. This can create an implicit language/locale expectation without user opt-in.
