The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its web content extraction functionality. * Ingestion points: Web content retrieved via the search.js --content and content.js tools. * Boundary markers: Absent in documentation; there are no instructions for the agent to treat the fetched content as untrusted or to use specific delimiters. * Capability inventory: The skill allows the agent to ingest arbitrary text from the web, which can influence subsequent reasoning steps. * Sanitization: Documentation does not indicate any filtering or escaping of the extracted markdown content.
[EXTERNAL_DOWNLOADS]: The skill's setup instructions utilize npm ci to download dependencies from the official npm registry, which is a well-known and trusted service.
[COMMAND_EXECUTION]: The skill requires the execution of local scripts (search.js and content.js) and the npm package manager to function.
[NO_CODE]: The core logic files (search.js and content.js) referenced in the documentation are missing from the provided skill files, which prevents a thorough security audit of the actual code behavior.

brave-search