Security audit
Security checks for vulnerabilities and agentic risk
This is a markdown-only planning skill for a specific llama.cpp endpoint feature request, with no executable code, credential handling, persistence, or hidden data movement observed.
Installers should be aware that the skill may activate for broad feature-request or remote-support language. Prefer explicit use by skill name or narrower trigger phrases if routing precision matters.
The skill description includes very broad trigger terms such as 'work-productivity', 'feature', 'request', 'support', and 'remote', which are common in many unrelated conversations. This can cause unintended invocation of the skill, leading the agent to apply specialized instructions in contexts where they do not belong and increasing the risk of scope confusion or prompt-routing mistakes.
The keyword trigger list is overly generic and contains short, everyday terms like 'feature', 'request', 'support', 'remote', and 'server'. In a routing system, these generic keywords can match many benign requests, causing accidental skill activation and misapplication of this skill's workflow to unrelated tasks.
The skill enables implicit invocation while advertising a very broad set of triggers such as 'work-productivity,' 'feature,' 'request,' 'support,' 'workflow,' 'artifact,' 'checklist,' 'analysis,' and 'implementation support.' This can cause the agent to invoke the skill in many unrelated contexts, exposing users to unintended prompt routing, over-collection of context, or manipulation if the skill content later becomes unsafe or overly influential.
The trigger phrase "Help me [Feature Request] Support for Remote llama.cpp Server via URL Endpoint" is highly generic because it begins with common conversational language ("Help me") and can match ordinary user requests outside the intended narrow skill scope. This can cause unintended activation of the skill, routing unrelated prompts into a specialized workflow and potentially producing misleading or inappropriate output for the user's actual task.
No suspicious patterns detected.
