Inspected source does not establish malicious behavior, but the default local code-execution tool has real sensitive-data exposure potential when exposed to an untrusted MCP caller. Treat it as a dangerous capability warning, not a malware block.
Static reason
High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; previous stored version diff introduced dangerous source
Trigger
An MCP client calls the execute tool with code.
Impact
A compromised or over-privileged MCP client could run code with access to process environment data and package files, and send data to the configured API host.
Mechanism
Local evaluation of MCP-supplied TypeScript in a permissioned Deno worker.
Evidence
package.jsonsrc/server.tssrc/code-tool.tssrc/code-tool-worker.tssrc/http.tssrc/auth.tssrc/options.tssrc/index.ts