Security audit
Security checks for vulnerabilities and agentic risk
This skill is a generic GitHub workflow helper with overly broad activation wording, but I found no hidden execution, persistence, credential handling, or destructive behavior.
Before installing, note that the skill may activate more often than intended because its triggers are generic. Use it for GitHub or repository workflow help, and review any proposed code, CLI, or API actions before allowing an agent to run them.
The trigger list is broad and includes generic terms like 'run', 'api', 'issues', and 'github', which can cause the skill to activate for many unrelated developer requests. Over-broad activation increases the chance of incorrect routing, unintended application of the skill, and prompt-surface expansion where this skill's instructions influence tasks outside its intended scope.
The trigger sentence begins with a very broad everyday phrase ('Help me'), which can cause the skill to activate in many unrelated conversations. Overbroad activation increases the chance of unintended invocation, context confusion, and accidental execution of workflows in situations the user did not intend.
The README presents a GitHub interaction workflow with action-oriented keywords like run, api, issues, and prs, but does not warn users that such workflows may modify repositories, create artifacts, or interact with remote services. Because the skill is positioned as practical workflow support for bug fixes and hardening, the absence of explicit safety boundaries makes accidental destructive or privacy-impacting actions more plausible.
The trigger examples are broad, repetitive, and partially truncated, which increases the chance the skill is invoked for loosely related requests without clear user intent. In a GitHub-interaction context, ambiguous activation can lead an agent to enter workflows involving repositories, issues, PRs, or CLI actions when the user only wanted general advice, raising the risk of unintended consequential operations.
The listed trigger phrases are extremely generic and include common help-seeking language, making accidental invocation or unintended routing likely. In an agent ecosystem, broad trigger matching can cause this skill to activate outside its intended scope, which may mis-handle user requests, override more appropriate skills, or expose users to unintended workflows.
The description says to use the skill whenever a user asks for broad categories like software-and-data, github, interact, cli, issue, or general workflow support, without clearly limiting when the skill should or should not run. This creates ambiguous invocation behavior and can lead to accidental selection for routine requests that do not need this skill, reducing predictability and potentially bypassing more appropriate, narrower skills.
The example trigger sentences use generic phrasing like 'Help me' and 'I need a practical workflow', which is common across many harmless requests and not unique to this skill's domain. These examples reinforce loose matching behavior and make accidental invocation more likely, especially in systems that learn or rank from trigger examples.
The description says the skill should be used when a user asks about broad categories like software-and-data, github, interact, cli, or issue, or whenever they need a practical workflow or analysis for the requirement. This boundary is vague and expansive, so the orchestrator may select the skill for many unrelated prompts, increasing accidental invocation and mis-scoped guidance.
The trigger keyword list includes very broad terms such as "run", "api", "issue", and "github", which can match many unrelated user requests and cause the skill to be invoked outside its intended scope. Over-broad activation increases the chance of unintended behavior, poor routing, and unsafe assistance being applied in the wrong context.
The default_prompt contains broad, everyday-language phrases like 'help me' and generic workflow terms, which can cause the skill to match or be invoked for loosely related requests. Because implicit invocation is enabled, this increases the chance of accidental routing into this skill when the user did not clearly intend it, potentially exposing the user to incorrect actions, confusing behavior, or unintended tool usage.
The trigger sentence is ambiguous and does not clearly define when the skill should be activated, instead embedding a long requirement summary that is not a precise user intent. This can lead to accidental routing of unrelated requests into this skill, reducing reliability and potentially exposing users to inappropriate actions or outputs for the wrong context.
This file is explicitly a Chinese-language README variant, but the content does not state that language selection is optional or explain any locale-specific requirement. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is documented and justified.
The example trigger phrases are generic templates that largely restate the demand statement rather than giving precise activation patterns. This can train users or upstream tooling toward ambiguous invocation behavior, reinforcing over-selection of the skill and reducing routing reliability.
No suspicious patterns detected.
