Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

Known CVEnpm/@anthropic-ai/claude-code

CVE-2025-54795: Claude Code echo command bypasses the approval prompt

High

A crafted echo command smuggles other commands past the approval check, so prompt-injected content can run code without the user confirming. Fixed in 1.0.20.

Affected and fixed versions

Affected: < 1.0.20. Fixed in 1.0.20.

Sources

See every known CVE, and the ones in the news, in the Learning Center