Attributed metadata history
Up to 100 retained publisher events per page for 8998fd1cd910bba18b05412eb8871d88f9fa4422295456d73ad7ff5125da290a. Withdrawals and unavailable references remain attributed to their publishers.
intelligence_src_c25d91fa3ad9ae9ac7
observed · 2026-10-12 14:53:04 UTC
Sandbox-escape-style attacks can happen when an AI is able to modify its own configuration settings, such as by writing to configuration files. That was the case with Amp, an agentic coding tool built by Sourcegraph. The AI coding agent could update its own configuration and: Allowlist bash commands or Add a malicious MCP server on the fly to run arbitrary code This could have been exploited by the model itself, or during an indirect prompt injection attack as we will demonstrate in this post.
Fingerprint: 9212295750c41ca2a850c202ca162079d3ac2cb84b834a06bc75d032090f48e4
intelligence_src_c25d91fa3ad9ae9ac7
observed · 2026-10-12 14:53:03 UTC
available
Fingerprint: 0a2f1bebc7ea800736eafab237a0ddecfb8797a02d23951e8aa80935e2f3446f