Agenta is an open-source LLMOps platform. In Agenta-API prior to version 0.48.1, a Python sandbox escape vulnerability existed in Agentas custom code evaluator. Agenta used RestrictedPython as a sandboxing mechanism for user-supplied evaluator code, but incorrectly whitelisted the numpy package as safe within the sandbox. This allowed authenticated users to bypass the sandbox and achieve arbitrary code execution on the API server. The escape path was through numpy.ma.core.inspect, which exposes Pythons introspection utilities — including sys.modules — thereby providing access to unfiltered system-level functionality like os.system. This vulnerability affects the Agenta self-hosted platform (API server), not the SDK when used as a standalone Python library. The custom code evaluator runs server-side within the API process. The issue is fixed in v0.48.1 by removing numpy from the sandbox allowlist. In later versions (v0.60+), the RestrictedPython sandbox was removed entirely and replaced with a different execution model.
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.
Affected Software
| Name | Vendor | Start Version | End Version |
|---|---|---|---|
| Agenta | Agentatech | * | 0.48.1 (excluding) |
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