Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward web research workflow with citations, with minor notes around broad activation and saving long reports locally.
Install if you want an agent to run web-heavy research and produce cited reports. Be aware it may activate broadly for research-like prompts and may save long reports locally, so avoid using it for sensitive topics unless you are comfortable with that output being written to disk.
The trigger phrases and description are broad enough to match many ordinary user requests, which can cause the skill to activate in situations where the user did not explicitly intend a deep web-research workflow. In an agent setting, over-broad activation increases the chance of unnecessary tool use, unexpected external data retrieval, and scope expansion beyond the user's intent.
The activation section allows invocation for nearly any topic asked 'in depth,' without domain boundaries, user-confirmation gates, or sensitivity checks. Because this skill performs multi-source search and fetch, ambiguous activation can lead to unintended browsing behavior, excessive data collection, or use in sensitive contexts without deliberate consent.
The workflow instructs the agent to save long reports to a file without mentioning user notice, confirmation, destination, or sensitivity checks. Implicit file creation can surprise users, persist sensitive research content locally, and create privacy or data-handling issues in environments where file writes are not expected.
No suspicious patterns detected.
