The Agent Skills Directory

[Indirect Prompt Injection] (LOW): This skill retrieves data from an external source (Brave Search API) which includes titles and descriptions of web entities. This creates a surface for indirect prompt injection if the agent processes these descriptions as instructions. However, the risk is minimal as the provider is a reputable search engine and the data is limited to short snippets.
[Data Exposure] (SAFE): The skill documentation correctly references the use of an environment variable (${BRAVE_SEARCH_API_KEY}) for authentication rather than hardcoding credentials.

suggest