CVE-2026-77177

Summary

The flaw allows code execution via prompt injection with Jinja2 template syntax, leading to server-side expression evaluation without proper sanitization.

Exploitability

Exploitation is relatively straightforward given the lack of sanitization, requiring an attacker to craft a malicious prompt.

Blast radius

If exploited, this could lead to full server compromise, potentially affecting Meta AI backend services for WhatsApp and other products.

Detection

No reliable host or network indicator is derivable from the published description.

Prioritized remediation

Disable the affected feature or upgrade to the latest version of Open GenAI Stack (ogx-ai) as soon as possible.

Analysis generated locally by qwen2.5:7b-instruct (no data left the box). AI-assisted — verify against primary sources before acting.