MediumVulnerabilityLLM-specific
CVE-2026-126767: OpsTree REMS prompt injection exposes hidden system prompt via /api/v1/rca/chat
- Source
- NVD(opens in a new tab)
- Identifier
- CVE-2026-126767
- Published
- Record updated
- Known exploitation
- Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
- EPSS
- 0.3%
Summary
OpsTree REMS through 1.0.0 contains a prompt injection vulnerability, tracked as CVE-2026-126767. Authenticated users can send crafted chat input to POST /api/v1/rca/chat to make the assistant ignore its rules and disclose its hidden system prompt, including internal orchestration instructions. That disclosure can help an attacker craft further prompt injection attacks.
Mitigation
The source does not state a fix yet. Check the original advisory for updates.
Topics
Related items
- LowYour Robot's Safety Functions Already Work. What If the Input Lies?Similar attack · VicOne Blog
- InfoTop GenAI security resources — November 2026Similar attack · Adversa AI Blog
- MediumAI SAST Finding: Path Traversal in OpenClaw via LLM Guardrail Bypass | Endor LabsSimilar attack · Endor Labs Blog
- InfoSecurity, Privacy, and Trustworthy Design for Retrieval-Augmented Generation SystemsSimilar attack · OpenAlex (peer-reviewed AI security)
- HighCVE-2026-63651: LiveDocs is a primary-source-first documentation for AI agents. Prior to version 0.7.0, fetch_docs and introspect…Similar attack · NVD/CVE Database