Anthropic has updated its Claude Managed Agents with new security features, introducing web domain allowlists and blocklists for its web_search and web_fetch tools. These lists, which can each contain up to 64 entries, provide granular control over the sites agents can access. A key distinction is that while web_fetch errors when a URL is not permitted, web_search silently omits disallowed results, making it harder to detect filtering. Crucially, these agent-specific lists override the broader domain restrictions set in the Claude Console, which only apply to Messages API requests. AI
IMPACT Enhances security and control for AI agents, potentially impacting how developers implement web browsing capabilities.
RANK_REASON This is a product update for an existing AI agent service, not a new frontier model release or significant industry-wide event.
Read on dev.to — Anthropic tag →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
