Anthropic response to 1-click pwn: Shouldn't have clicked 'ok'

The TrustFall proof-of-concept attack demonstrates how a cloned code repository can include two JSON files (.mcp.json and ...

200,000 MCP servers expose a command execution flaw that Anthropic calls a feature

OX Security confirmed arbitrary command execution on six live platforms and estimates 200,000 MCP servers are exposed. Here's ...
Hosted on MSN

Vibe coding tool Cursor's MCP implementation allows persistent code execution

Check Point researchers uncovered a remote code execution bug in popular vibe-coding AI tool Cursor that could allow an attacker to poison developer environments by secretly modifying a previously ...

Running Claude Code or Claude in Chrome? Here's the audit matrix for every blind spot your security stack misses

Four research teams found the same confused deputy failure in Claude across three surfaces in 48 hours. This audit matrix ...
CSOonline

Rogue MCP servers can take over Cursor’s built-in browser

A new proof-of-concept attack shows that malicious Model Context Protocol servers can inject JavaScript into Cursor’s browser — and potentially leverage the IDE’s privileges to perform system tasks.

DataDoe Launches Amazon Data MCP for Claude, ChatGPT and Cursor

DataDoe gives AI tools live Seller Central, Ads, inventory, fees and profit data so ecommerce teams can build on real ...