Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation-only developer workflow helper with overly broad activation wording, but no hidden code, persistence, credential handling, or automatic external actions.
This skill appears safe to install as a general developer workflow helper, but expect it to activate too broadly unless the platform or publisher narrows the trigger terms or disables implicit invocation. Review its suggestions before allowing any separate tool to modify repositories, issues, or pull requests.
The trigger list is broad enough to match common developer requests such as 'github', 'cli', 'api', 'issue', and 'run', which can cause the skill to activate when the user did not intend to invoke it. In an agent environment, unintended activation can route tasks through the wrong workflow, leading to inappropriate actions, misleading outputs, or bypass of more suitable safeguards in other skills.
The trigger sentence begins with an extremely broad conversational phrase ('Help me ...'), which can cause the skill to activate for many ordinary user requests that are not actually about this specific GitHub-style workflow. In an agent-routing context, overbroad triggers can misroute tasks, expand the skill's effective authority, and make it easier for unrelated or adversarial prompts to invoke the skill unintentionally.
The trigger sentences are so generic and truncated that they can match a wide range of unrelated requests, causing the skill to be invoked outside its intended scope. In an agent ecosystem, overly broad invocation rules can lead to unintended execution paths, incorrect workflow selection, and increased exposure to downstream risky actions such as repository interaction or issue/PR handling without strong user intent.
The invocation section lists generic trigger patterns like "Help me" and "I need a practical workflow" wrapped around a long descriptive sentence, without clear scope boundaries or negative examples. These phrases overlap with common everyday language and do not precisely define when the skill should activate versus when it should not.
The example trigger sentences use vague, everyday help-request phrasing that does not clearly distinguish this skill from many unrelated software-support tasks. This increases the chance of accidental invocation and compounds the ambiguity created by the broad keyword list, especially in systems that learn or match from examples.
The trigger keyword list is overly broad and includes generic terms like 'run' and 'api' alongside common software terms. This can cause the skill to activate in unrelated contexts, leading to unintended routing, over-collection of user context, or inappropriate workflow guidance being injected into conversations that did not request this skill.
The example trigger phrases use highly generalized help-seeking language such as 'Help me' and 'I need a practical workflow,' which weakens activation boundaries. In systems that learn or match from examples, this can broaden invocation behavior and cause accidental skill selection for ordinary support requests unrelated to GitHub-style developer workflows.
The default prompt/activation text is extremely broad and matches generic software help-seeking requests such as fixing bugs, improving safety, reliability, workflows, checklists, and implementation support. Because implicit invocation is enabled, this creates a realistic risk that the skill is triggered in contexts beyond its intended scope, causing unreviewed prompt injection, unintended tool usage, or unexpected handling of sensitive repository/workflow tasks.
The trigger scope is ambiguous because it mixes generic verbs ('help', 'need') with a long requirement statement rather than crisp activation constraints. This makes the skill easier to invoke accidentally across unrelated software tasks, which is especially risky for a broad 'developer helper' skill that may influence code, workflows, or repository operations.
This file is entirely presented in Simplified Chinese and does not state that language selection is optional or that the skill is region-specific. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is clearly justified.
No suspicious patterns detected.
