A critical remote code execution (RCE) vulnerability has been discovered in Python PLY (Python Lex-Yacc), a well-known parsing library that has been part of the Python ecosystem for nearly 25 years. The flaw, identified just as the project’s creator officially announced the library’s abandonment, exposes applications to full system compromise if exploited.
The vulnerability is tracked as CVE-2025-56005 and has been assigned a CVSS score of 9.8, placing it in the highest severity category. Security researchers warn that any application using vulnerable versions of PLY could be at serious risk, especially in environments where user input can influence parser behavior.
PLY (Python Lex-Yacc) is a popular zero-dependency parsing library that implements the traditional lex and yacc tools in pure Python. Developers have long used PLY to build custom parsers, compilers, interpreters, and domain-specific languages.
Because of its simplicity and compatibility with modern Python versions, PLY has been included in many legacy applications, internal tools, academic projects, and even production systems. Despite its age, many developers continued to trust PLY due to its stability and long-standing reputation.
However, that trust is now being questioned.
The vulnerability was discovered by security researcher Bohmiiidd, who found a hidden and undocumented parameter inside the yacc() function of the PLY library. This parameter, named picklefile, exists in the released PyPI package but is not mentioned anywhere in the official documentation or GitHub repository.
According to the researcher’s findings:
“This parameter is not documented in the official PLY documentation or GitHub repository, yet it is active in the PyPI release.”
Because the feature is undocumented, most developers are completely unaware that it exists.
The picklefile parameter appears to have been designed for caching parser state or saving internal data. However, its implementation introduces a serious security risk.
When this parameter is used, PLY automatically deserializes a file using Python’s built-in pickle.load() function without any validation or security checks.
This is extremely dangerous.
In Python, the pickle module is not safe for untrusted data. During deserialization, pickle can execute arbitrary Python code using mechanisms like the __reduce__() method.
This means:
If an attacker can control the pickle file
Or influence the value passed to
picklefileThey can execute any system command on the target machine
In simple terms, this flaw allows remote code execution the moment the parser is initialized.
To demonstrate the severity of the issue, Bohmiiidd released a proof of concept (PoC) showing how a malicious pickle payload can be used to execute system-level commands.
The PoC shows that as soon as the yacc() function is called with a crafted pickle file, the attacker’s code runs immediately. This could allow attackers to:
Create or delete files
Steal sensitive data
Install malware or backdoors
Take complete control of the server
In production environments, this kind of vulnerability could lead to full infrastructure compromise.
What makes this situation even more serious is the timing of the disclosure.
The creator of PLY, David Beazley, recently announced that he has officially abandoned the project after maintaining it for over two decades. In his announcement, he made it clear that:
-
No further development is planned
-
No security patches will be released
-
No official fixes are coming
Beazley advised developers to move away from PLY and adopt modern parsing libraries instead, stating:
“At this point, there are many high-quality parsing libraries that you might consider using instead.”
This means organizations cannot rely on an upstream fix and must take immediate action themselves.
Applications are most at risk if:
-
They use Python PLY in production
-
They dynamically generate parsers
-
User input can influence arguments passed to
yacc() -
The application loads external or user-controlled files
Legacy systems and internal tools are especially vulnerable, as they often lack modern security reviews.
Since there is no official patch, mitigation is critical.
-
Do not use the
picklefileparameter
The feature is undocumented and should never be used. -
Audit your codebase
Search for any usage ofyacc()and verify that no untrusted input can influence its parameters. -
Restrict user input
Ensure that attackers cannot control file paths or function arguments related to parser initialization. -
Plan migration to safer alternatives
Consider actively maintained parsing libraries with modern security practices.
The discovery of CVE-2025-56005 highlights a major risk associated with abandoned open-source software. Even trusted libraries can hide dangerous behavior, especially when undocumented features are involved.
If your organization still relies on Python PLY, this vulnerability should be treated as high priority. Without patches or maintenance, the safest long-term option is to move away from the library entirely.
In cybersecurity, old code never truly disappears and sometimes, it comes back as a ghost in the system.
Interesting Article : Critical VMware vCenter RCE Flaw CVE-2024-37079 Under Active Attack
Pingback: Microsoft Office Zero-Day CVE-2026-21509 Actively Exploited