Security audit
Security checks for vulnerabilities and agentic risk
This is a straightforward API-reference skill for using Presearch web search, with expected external search requests and no executable local behavior.
Install only if you are comfortable sending search terms to Presearch's API and managing a Presearch bearer token. Do not include passwords, secrets, private customer data, or confidential project details in search queries.
The skill documentation describes a search capability but does not clearly warn that user queries are sent to an external third-party endpoint at presearch.com. This can expose sensitive prompts, internal project names, credentials accidentally included in queries, or other private data to an outside service, creating a real privacy and data-governance risk for agents and users.
Defaulting the language to en-US without clear user choice can bias results, reduce accuracy for non-English users, and cause inadvertent disclosure of locale assumptions in agent behavior. In a search skill this is not a direct exploit primitive, but it is a legitimate security/privacy quality issue because hidden defaults can mis-handle user intent and route sensitive searches under an unintended locale context.
No suspicious patterns detected.
