None — every tool is low-risk, high-confidence, and outside PCI/HIPAA payload scope.
All other tools (5)
tavily.tavily_crawl medium conf highaudit
Recursively harvests content across a website at configurable depth and breadth, a bulk scraping operation that ingests large volumes of third-party data and can breach a site's terms of service.
tavily.tavily_extract low conf highallow
Pulls the full raw content of specified URLs into the agent context, potentially ingesting third-party personal data or injected instructions from those pages.
tavily.tavily_map low conf highallow
Enumerates the URL structure of a target website, exposing its page inventory without fetching page bodies.
tavily.tavily_research medium conf mediumaudit
Runs multi-source aggregated research and returns a synthesized answer, blending many untrusted web sources into one response and compounding provenance risk.
tavily.tavily_search low conf highallow
Retrieves live web search snippets and source URLs, pulling untrusted external content into the agent context that can steer downstream decisions.