CVE-2026-109440

41th percentile - higher than 41% of all known CVEs

LangAlpha through 2026.09.07 contains an indirect prompt injection vulnerability that allows remote attackers to execute shell commands or Python by planting instructions in web content retrieved by the agent. Because Bash and ExecuteCode lack an allow-list and human-in-the-loop approval, attackers can run code in the workspace sandbox and exfiltrate vault secrets from _internal/.vault_secrets.json.

Arbitrary code execution in the workspace and leakage of vault secrets are possible, potentially leading to further escalation and data compromise.

Update LangAlpha to a version later than 2026.09.07; until then, enforce an allow-list and human-in-the-loop approval for Bash and ExecuteCode tools.