Weakness (CWE)
CWE-95: Improper Neutralization of Directives in Dynamically Evaluated Code
The product receives input from an upstream component, but it does not neutralize or incorrectly neutralizes code syntax before using the input in a dynamic evaluation call (e.g. "eval").
How to prevent it (4 mitigations)
-
Architecture and Design, Implementation
If possible, refactor your code so that it does not need to use eval() at all.
-
Implementation
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.
-
Implementation
Inputs should be decoded and canonicalized to the application's current internal representation before being validated (CWE-180, CWE-181). Make sure that your application does not inadvertently decode the same input twice (CWE-174). Such errors could be used to bypass allowlist schemes by introducing dangerous inputs after they have been checked. Use libraries such as the OWASP ESAPI Canonicalization control.
Consider performing repeated canonicalization until your input does not change any more. This will avoid double-decoding and similar scenarios, but it might inadvertently modify inputs that are allowed to contain properly-encoded dangerous content.
-
Implementation
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].
Source: MITRE CWE-95
116 results · Clear filters · Feed
-
Critical 9.4
GroupOffice Vulnerable to RCE in TemplateParser eval() via Contact/labels
Group-Office is an enterprise customer relationship management and groupware tool. Prior to versions 6.8.167, 25.0.102, and 26.0.27, the Contact/labels JMAP endpoint in GroupOffice accepts a fully attacker-controlled template string (tpl parameter) that is pa…
-
Critical 9.5
[Eclipse CSI - otterdog] Unauthenticated remote code execution via eval() on format-string-injected webhook body when GITHUB_WEBHOOK_SECRET is unset
The otterdog webapp evaluates a format-string built from the incoming GitHub webhook JSON body using Python's eval(). When the GITHUB_WEBHOOK_SECRET configuration value is unset or empty (the code default is None), the HMAC signature check is skipped entirely…
-
Critical 9.2
Code execution through double evaluation in address replacement rules
SEPPmail Secure Email Gateway before 15.1.1 can evaluate attacker-influenced replacement text as Perl code in sender or recipient replacement rules. An unauthenticated remote sender can execute operating-system commands in the mail-processing worker context w…
-
Critical 9.8
Spreadsheet::XLSX versions before 0.19 for Perl allow code execution via string eval of number format conditions in ExcelFmt
Spreadsheet::XLSX versions before 0.19 for Perl allow code execution via string eval of number format conditions in ExcelFmt. When a number format has a condition such as [>100], ExcelFmt chooses the format section by passing the condition and the cell value …
-
Critical 9.9
XWiki Pro Macros: excerpt-include title and content injection executes code
XWiki Remote Macros provides XWiki rendering macros that are useful when migrating content from Confluence. From 1.13 until 1.14.5, the `excerpt-include` macro does not properly escape the included page's title and executes `excerpt` content with the macro's …
-
Critical 9.8
WeRSS (we-mp-rss) through 1.5.3 RCE via template Parameter in /rss Endpoint
WeRSS (we-mp-rss) through 1.5.3 contains a remote code execution vulnerability that allows unauthenticated attackers to run code via the template parameter of the public GET /rss/{feed_id} endpoint. Attackers can supply a non-built-in ext value and a {{=expre…
-
Critical 9.0
Dokku: Host Command Execution via eval of Unsanitized Docker Options in DOCKER_ARGS
Dokku is a docker-powered platform as a service. In Dokku versions before 0.38.25, a user with app-level Dokku access and remote SSH access to execute docker-options commands can store arbitrary strings through the docker-options plugin, and those strings are…
-
Critical 9.6
Affected devices do not properly restrict executable content contained in imported trace files. This could allow an att…
-
Critical 9.9
Astron Agent: Unsandboxed code-node leads to cross-tenant RCE
Astron Agent is an agentic workflow platform for building and running AI agents. Prior to 1.1.2, the default workflow code-node path through /console-api/workflow/code/run and /workflow/v1/run selects LocalExecutor in core/workflow/engine/nodes/code/code_node…
-
Critical 9.8
Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in hMailServer
Eval injection in the JScript event-script dispatcher in Progressive Robot Ltd's hMailServer, versions 6.0.0 through 6.3.3 on Windows, allows a remote, unauthenticated attacker to run arbitrary JScript inside the hMailServer service process, with the privileg…
-
Critical 9.9
plone.app.portlets Vulnerable to Remote Code Execution via TALES Injection
plone.app.portlets.portlets provides a Plone-specific user interface for plone.portlets, as well as a standard set of portlets that ship with Plone. Starting in version 5.0.0 and prior to versions 5.0.8, 6.0.4, and 7.0.2, the Classic portlet (plone.app.portle…
-
Critical 9.8
An issue in gray-matter All versions (verified on 4.0.3) allows the JavaScript engine in lib/engines.js using eval() to…
-
Critical 9.9
org.xwiki.rendering:xwiki-rendering-xml has an Eval Injection issue
XWiki Rendering is a generic rendering system that converts textual input in a given syntax (wiki syntax, HTML, etc) into another syntax (XHTML, etc). Prior to versions 14.10.2 and 15.0 RC1, any user who can edit their own user profile or any other document c…
-
Critical 9.9
DIRAC: RCE in FileCatalog DatasetManager via SQL injection + eval
DIRAC is an interware, meaning a software framework for distributed computing. Prior to versions 8.0.79, 9.0.22, and 9.1.10, DataManagementSystem/Service/FileCatalogHandler.py checkDataset forwards an authenticated caller-controlled datasets value to DatasetM…
-
Critical 9.9
DIRAC: RCE in RequestManager due to eval on untrusted input
DIRAC is an interware, meaning a software framework for distributed computing. Prior to versions 8.0.79, 9.0.22, and 9.1.10, the RequestManagementSystem/Service/ReqManagerHandler.py export_getRequestCountersWeb function passes an authenticated caller-controll…
-
Critical 9.8
Apache Camel K: Camel K Tenant repositories reach Maven execution inside operator pod
Improper neutralization of directives in dynamically evaluated code ('eval injection') vulnerability in Apache Camel K. An improper neutralization of directives in dynamically evaluated Maven configuration allows tenant-controlled repository content to influe…
-
Critical 9.9
ColdFusion | Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') (CWE-95)
ColdFusion is affected by an Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') vulnerability that could result in arbitrary code execution in the context of the current user. A low-privileged attacker could exploit this vu…
-
Critical 9.9CVE-2026-19295 Public exploit
Langflow is affected by multiple remote code execution vulnerabilities due to insufficient code-execution policy enforcement
IBM Langflow OSS 1.0.0 through 1.11.1 allows an authenticated attacker to execute arbitrary operating system commands in the server process by saving a flow with a crafted type field value and triggering a build of a wrapper flow that references it. This allo…
-
Critical 9.8
SENAITE.CORE: Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') and Missing Authorization in senaite.core
SENAITE.CORE is the core framework for the SENAITE laboratory information management system. From 2.0.0 to 2.6.0, the SENAITE.CORE JSON API permits unauthenticated remote code execution through a two-request chain involving missing authorization and unsafe ev…
-
Critical 9.2
Eval Injection in google/langfun via default lf.query protocol
Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2 allows remote unauthenticated attackers to execute arbitrary Python code in the context of…
-
Critical 10.0
Xinference: Remote code execution via unsafe `eval()` in Llama3 tool-call parsing
Xinference is an inference API for running open-source, speech, and multimodal models. In 2.5.0 and earlier, Xinference passes attacker-influenced Llama3 tool-call output to eval() in xinference/model/llm/tool_parsers/llama3_tool_parser.py and xinference/mode…
-
Critical 9.9
Code Injection via Gremlin Query Passthrough in Amazon Athena Neptune Connector
In the Neptune connector, a user with access to Neptune through Athena Federated Query could gain access to properties in the Lambda supplying the compute for the connector. To remediate this issue, users should upgrade to aws-athena-query-federation v2026.30…
-
Critical 9.3
Orval: RCE via schema property name -> computed-property-key injection in the MSW mock generator
Orval generates type-safe JavaScript clients in TypeScript from OpenAPI v3 and Swagger v2 specifications. Prior to 8.21.0, a single quote in a schema property name is emitted into single-quoted object keys in generated MSW mock factories without safe encoding…
-
Critical 9.3
Orval: Import-time RCE via query parameter name -> computed-property-key injection in the zod cli
Orval generates type-safe JavaScript clients in TypeScript from OpenAPI v3 and Swagger v2 specifications. Prior to 8.21.0, a double quote in a query parameter name is emitted into the generated request-validation zod.object({...}) schema without safe encoding…
-
Critical 9.3
Orval: Import-time RCE via header parameter name -> computed-property-key injection in the zod client
Orval generates type-safe JavaScript clients in TypeScript from OpenAPI v3 and Swagger v2 specifications. Prior to 8.21.0, a double quote in a header parameter name is emitted into the generated request-validation zod.object({...}) schema without safe encodin…
-
Critical 9.3
Orval: Import-time RCE via schema property name -> computed-property-key injection in the zod client
Orval generates type-safe JavaScript clients in TypeScript from OpenAPI v3 and Swagger v2 specifications. From version 8.19.0 until 8.21.0, a double quote in a schema property name is emitted into the generated zod.object({...}) schema without safe encoding. …
-
Critical 9.8
openssl_encrypt before 1.4.0 Sandbox Escape via Type Hierarchy
openssl_encrypt versions before 1.4.0 contain a sandbox escape vulnerability in IsolatedPluginExecutor that exposes Python type objects in restricted exec() builtins. Attackers can traverse the Python class hierarchy via __class__.__mro__.__subclasses__() to …
-
Critical 9.9CVE-2026-19626 Public exploit
Remote Code Execution
A remote code execution vulnerability exists in Tenable Security Center's report generation functionality. An authenticated, non-administrative user could exploit this issue by supplying specially crafted input that is later processed unsafely during server-s…
-
Critical 9.0
Flowise before 3.1.3 Sandbox Escape to RCE
Flowise before 3.1.3 contains a sandbox escape vulnerability in the vm2 JavaScript sandbox that allows authenticated users to execute arbitrary code by exploiting moment locale validation bypass. Attackers can craft a fake String object with a match function …
-
Critical 9.0
Flowise before 3.1.3 Remote Code Execution via Custom MCP
Flowise versions before 3.1.3 contain a remote code execution vulnerability in the Custom MCP node when CUSTOM_MCP_PROTOCOL is set to stdio, allowing authenticated users to execute arbitrary commands by manipulating environment variables and command arguments…
-
Critical 9.3
Fooocus webui vulnerable to Remote Code Execution
Fooocus is an image generating software. In versions 2.5.5 and prior, the Fooocus web UI is vulnerable to remote code execution due to the unsafe use of eval when processing metadata JSON. An attacker with access to the Fooocus web UI may be able to execute a…
-
Critical 9.3
Firecrawl: Arbitrary file read via JSON Schema $ref expansion
Firecrawl turns entire websites into LLM-ready markdown or structured data. Prior to 2.11.32, a critical arbitrary file read vulnerability exists in Firecrawl's extraction functionality due to unsafe schema dereferencing of user-supplied JSON schemas in apps/…
-
Critical 9.4
Flowise: RCE via CSVAgent csvFile data URI base64 segment is interpolated into Python source without validation
Prior to 3.1.3, Flowise CSVAgent interpolates an attacker-controlled segment of the csvFile data URI directly into a Python source-code template that is then executed by Pyodide. Because Pyodide is loaded with the default js bridge to globalThis, which on Nod…
-
Critical 9.0
Flowise Sandbox Escape to RCE
Flowise is a drag-and-drop user interface for building customized large language model (LLM) flows. Prior to version 3.1.3, several custom-tool components — AgentAsTool, ChatflowTool, and ExecuteFlow — ran code in the in-process vm2 sandbox. To build that c…
-
Critical 9.6
Adobe Campaign Classic (ACC) | Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') (CWE-95)
Adobe Campaign Classic (ACC) is affected by an Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') vulnerability that could result in arbitrary code execution in the context of the current user. A low-privileged attacker cou…
-
Critical 9.1
OpenEMR 8.2.0 Remote Code Execution via CategoryTree eval() Injection
OpenEMR through 8.2.0 contains a remote code execution vulnerability in the document category tree component (library/classes/Tree.class.php) that allows authenticated administrators to execute arbitrary operating system commands by injecting PHP payloads int…
-
Critical 9.8
CVE-2026-15971
SGLang contains an RCE vulnerability when the optional dumper subsystem is enabled, allowing for a sandbox escape when DUMPER_SERVER_PORT is set, enabling code execution on inference requests.
-
Critical 9.8CVE-2026-61511 Public exploit
vBulletin < 6.2.2 Eval Injection RCE via vb5/template/runtime.php
vBulletin 5.x through 5.7.5 and 6.x through 6.2.1 contains an eval injection vulnerability in the vB5_Template_Runtime::runMaths() method within the template runtime that allows unauthenticated remote attackers to execute arbitrary PHP code by supplying craft…
-
Critical 9.8
PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution
PraisonAI is a multi-agent teams system. Prior to version 4.6.40, PraisonAI's first-party A2A server example exposes an unauthenticated A2A JSON-RPC endpoint and registers a `calculate(expression)` tool implemented with Python `eval()`. The example also binds…
-
Critical 9.8
Net::DNS versions through 1.55 for Perl allow remote execution injection via EDNS EXTENDED ERROR
Net::DNS versions through 1.55 for Perl allow remote execution injection via EDNS EXTENDED ERROR. Net::DNS::RR::OPT::EXTENDED_ERROR::_decompose parses the EXTRA-TEXT field of an EDNS EXTENDED-ERROR option (RFC 8914) by tokenising the raw bytes and passing the…