The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external web search results, which presents a surface for indirect prompt injection.
Ingestion points: Search result context and AI-synthesized responses are retrieved from the api.search.brave.com API endpoint as described in SKILL.md.
Boundary markers: The response format utilizes structural tags including <answer>, <citation>, and <thinking> to separate different data types.
Capability inventory: The skill is designed for network-based information retrieval and does not involve local file system writes or unauthorized command execution.
Sanitization: The skill supports a safesearch parameter to provide content filtering at the source API.- [SAFE]: The skill exclusively interacts with official Brave infrastructure and provides clear documentation for secure API key management using environment variables.

answers