CVE-2026-22686
Published: 14 January 2026
Summary
CVE-2026-22686 is a critical-severity Code Injection (CWE-94) vulnerability in Agentfront Enclave. Its CVSS base score is 10.0 (Critical).
Operationally, exploitation aligns with the MITRE ATT&CK technique Exploitation for Privilege Escalation (T1068); ranked in the top 49.1% of CVEs by exploit likelihood; it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as Other AI Platforms.
The strongest mitigations our analysis identified are NIST 800-53 SC-50 (Software-enforced Separation and Policy Enforcement) and SI-11 (Error Handling).
Threat & Defense at a Glance
Threat & Defense Details
Mitigating Controls (NIST 800-53 r5)AI
Directly mitigates the sandbox escape by requiring identification, reporting, and timely patching of the specific flaw in enclave-vm prior to version 2.7.0.
Prevents exploitation by ensuring error handling during failed tool invocations does not expose host-side Error objects or exploitable prototype chains to sandboxed code.
Enforces software-based separation and policy between untrusted sandboxed JavaScript and the host Node.js runtime, blocking prototype chain traversal to the host Function constructor.
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
The vulnerability is a sandbox escape that exploits exposure of a host Error object to achieve arbitrary code execution on the host Node.js runtime, directly facilitating exploitation for privilege escalation (T1068) and defense evasion by bypassing sandbox isolation (T1211).
NVD Description
Enclave is a secure JavaScript sandbox designed for safe AI agent code execution. Prior to 2.7.0, there is a critical sandbox escape vulnerability in enclave-vm that allows untrusted, sandboxed JavaScript code to execute arbitrary code in the host Node.js runtime.…
more
When a tool invocation fails, enclave-vm exposes a host-side Error object to sandboxed code. This Error object retains its host realm prototype chain, which can be traversed to reach the host Function constructor. An attacker can intentionally trigger a host error, then climb the prototype chain. Using the host Function constructor, arbitrary JavaScript can be compiled and executed in the host context, fully bypassing the sandbox and granting access to sensitive resources such as process.env, filesystem, and network. This breaks enclave-vm’s core security guarantee of isolating untrusted code. This vulnerability is fixed in 2.7.0.
Deeper analysisAI
CVE-2026-22686 is a critical sandbox escape vulnerability (CVSS 10.0) in the enclave-vm component of Enclave, a secure JavaScript sandbox designed for safe AI agent code execution. Affecting versions prior to 2.7.0 and running on Node.js, the flaw (CWE-94, CWE-693) occurs when a tool invocation fails, causing enclave-vm to expose a host-side Error object to sandboxed code. This Error object retains its host realm prototype chain, enabling traversal to the host Function constructor and allowing untrusted JavaScript to compile and execute arbitrary code in the host Node.js runtime.
The vulnerability can be exploited by any untrusted, sandboxed JavaScript code executed within Enclave, such as code from AI agents. An attacker intentionally triggers a host error during a failed tool invocation, then climbs the exposed Error object's prototype chain to access the host Function constructor. This achieves full sandbox bypass, granting access to sensitive host resources including process.env, the filesystem, and network capabilities, thereby breaking Enclave's isolation guarantees. The CVSS vector (AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:H) indicates remote exploitation with no privileges or user interaction required.
The vulnerability is fixed in Enclave version 2.7.0. The GitHub security advisory (GHSA-7qm7-455j-5p63) and fixing commit (ed8bc438b2cd6e6f0b5f2de321e5be6f0169b5a1) detail the patch, which security practitioners should apply immediately to deployments using Enclave for isolating untrusted AI agent code.
Details
- CWE(s)
Affected Products
AI Security AnalysisAI
- AI Category
- Other AI Platforms
- Risk Domain
- N/A
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: ai