Security audit
Security checks for vulnerabilities and agentic risk
This skill is a documentation-only workflow helper, but its broad automatic trigger wording could make it activate for unrelated requests.
Installing this skill is reasonable if you want a lightweight PollyReach-style workflow helper. Review or narrow its trigger terms before relying on implicit invocation, because common words may cause it to appear for unrelated productivity requests.
The trigger keywords and example phrases are extremely generic, including common words like 'every', 'phone', 'number', 'ability', and 'things', plus broad helper-style sentences. This creates a high risk of accidental invocation or routing hijack, where ordinary user requests unrelated to this skill could match and activate it, leading to inappropriate execution paths or unexpected data handling.
The trigger guidance includes extremely broad everyday terms and a generic sentence pattern, which can cause the skill to activate for unrelated user requests. In an agent ecosystem, unintended invocation can misroute tasks, expose irrelevant workflow behavior, and increase the chance of unsafe or confusing automation.
The invocation description says to use the skill when a user asks for broad categories like "work-productivity" or needs any "workflow, artifact, checklist, analysis, or implementation support," which is far too open-ended. This can make the orchestrator select the skill for a wide range of unrelated requests, creating unsafe overreach, misrouting, and possible interference with more appropriate or restrictive skills.
The trigger keyword list is excessively broad and includes common everyday terms such as "every," "phone," "number," "ability," and "things," which can cause the skill to activate in many unrelated conversations. Over-broad activation increases the chance the agent routes user requests into this skill unintentionally, leading to prompt/skill hijacking of normal workflows and reducing reliability and safety of skill selection.
The trigger keyword list is overly broad and includes generic terms like "every", "phone", "number", "ability", and "done", which can cause the skill to activate for unrelated user requests. In an agent environment, accidental invocation can misroute tasks, override more appropriate skills, and create unsafe or confusing workflow execution paths.
The trigger examples include very generic everyday wording such as requests for a 'practical workflow,' which can match many unrelated user prompts and cause accidental skill invocation. Over-broad activation increases the chance that this skill is selected outside its intended PollyReach/work-productivity context, potentially leading to irrelevant actions, user confusion, or unsafe delegation in broader workflows.
A file labeled as Chinese (README.zh-CN.md) contains substantial English headings, descriptions, and trigger text, which can impose a language experience inconsistent with the locale-specific filename. There is no indication that users may choose their preferred language or that mixed-language content is intentional.
The invocation documentation is ambiguous and overly permissive, making it unclear when the skill should be selected versus other skills. This creates prompt-routing risk: unrelated conversations may invoke the skill, producing incorrect actions or outputs and reducing user control over agent behavior.
The example trigger sentences are generic templates like "Help me" and "I need a practical workflow," followed by truncated requirement text, so they do not meaningfully constrain activation. Such examples can train routing logic or maintainers toward permissive matching behavior, increasing accidental invocation and weakening separation between skills.
The skill description says it should be used when users mention broad terms like work-productivity, pollyreach, gives, every, or phone, but it does not define clear eligibility conditions or exclusions. This ambiguity increases the chance that the orchestration layer will select the skill in contexts outside its intended purpose, reducing predictability and potentially causing incorrect task handling.
The example trigger phrases are highly generic and resemble normal conversational requests, so they do not meaningfully constrain when the skill should be called. These examples can train routing logic or human operators toward over-broad activation, increasing accidental selection and workflow confusion.
The skill enables implicit invocation while using a very broad description and default prompt tied to generic productivity terms, which can cause the agent to trigger this skill in situations the user did not intend. That increases the risk of prompt-routing errors, unintended disclosure of user context to the skill, and unreliable or unsafe workflow execution, especially because the skill appears applicable to many common requests.
The trigger guidance is ambiguous and underspecified, mixing broad keywords ('gives,' 'every,' 'phone') with incomplete trigger sentences that do not clearly delimit when the skill should or should not activate. In an agent ecosystem, ambiguous routing logic can misfire frequently, causing unintended invocation and widening the attack surface for prompt-routing abuse or operational mistakes.
No suspicious patterns detected.
