Amp Code: Arbitrary Command Execution via Prompt Injection Fixed
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 .
Summary from Embrace The Red.
Explained
The technology and terms in this story. Learn more
- AttackPrompt injectionHiding instructions in content an AI reads, so it follows the attacker instead of its user.
- TechnologyAI agentAn AI model that acts on its own: it plans steps, uses tools and takes actions, not just answers.
- ProtocolMCP · Model Context ProtocolA standard way to connect AI models to tools and data sources.
- ConceptExploitCode or a technique that turns a software flaw into something an attacker can actually use.