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. 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.
Weakness
The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.
Potential Mitigations
- Run your code in a “jail” or similar sandbox environment that enforces strict boundaries between the process and the operating system. This may effectively restrict which code can be executed by your product.
- Examples include the Unix chroot jail and AppArmor. In general, managed code may provide some protection.
- This may not be a feasible solution, and it only limits the impact to the operating system; the rest of your application may still be subject to compromise.
- Be careful to avoid CWE-243 and other weaknesses related to jails.
- Assume all input is malicious. Use an “accept known good” input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does.
- When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, “boat” may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as “red” or “blue.”
- Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code’s environment changes. This can give attackers enough room to bypass the intended validation. However, denylists can be useful for detecting potential attacks or determining which inputs are so malformed that they should be rejected outright.
- To reduce the likelihood of code injection, use stringent allowlists that limit which constructs are allowed. If you are dynamically constructing code that invokes a function, then verifying that the input is alphanumeric might be insufficient. An attacker might still be able to reference a dangerous function that you did not intend to allow, such as system(), exec(), or exit().
- For Python programs, it is frequently encouraged to use the ast.literal_eval() function instead of eval, since it is intentionally designed to avoid executing code. However, an adversary could still cause excessive memory or stack consumption via deeply nested structures [REF-1372], so the python documentation discourages use of ast.literal_eval() on untrusted data [REF-1373].
Related Attack Patterns
- https://cwe.mitre.org/data/definitions/242.html
- https://cwe.mitre.org/data/definitions/35.html
- https://cwe.mitre.org/data/definitions/77.html