Skip to main content
Mallory
CriticalExploited in the wildCISA KEVPublic exploitVulnerability

CVE-2024-4577: PHP-CGI Argument Injection on Windows

Published Jun 9, 2024 · Updated 4 days ago
Trending this weekNot trendingOutside this week’s top 100
Mention activity · 30 days25 mentions · peak 15 in a dayToday
CVSS9.8Critical · v3.1
EPSS100.0%100th percentile
CISA KEVYesAdded Jun 12, 2024
In the wildActivePer CISA KEV
Vendors
Fedora ProjectPHP Group
IdentifiersCVE-2024-4577CWE-88

CVE-2024-4577 is an argument-injection vulnerability in PHP-CGI running with Apache on Windows under certain code-page configurations. Windows “Best-Fit” character conversion can transform attacker-controlled characters in command lines passed to Win32 API functions into characters that PHP-CGI interprets as command-line options. Unauthenticated remote attackers can submit specially crafted URL query arguments to inject PHP options, potentially disclosing script source code or executing arbitrary PHP code. Affected releases include PHP 8.1.x before 8.1.29, 8.2.x before 8.2.20, and 8.3.x before 8.3.8; unsupported older branches may also be affected. The vulnerability has been exploited in the wild, including in ransomware campaigns, and is listed in CISA’s Known Exploited Vulnerabilities catalog.

Do you run this?See affected versions in your environment and whether adversaries are exploiting this vulnerability against assets like yours.
Check your exposure
Analyst brief

Impact, mitigation & remediation

What it means and what to do now: patch path, interim mitigations, and the assume-compromise checklist.

Impact

What an attacker gets, and what they’ve been doing with it.

Successful exploitation can compromise application confidentiality, integrity, and availability through script source-code disclosure and arbitrary code execution in the affected service account’s security context. Attackers may read application secrets, modify or delete accessible data, and install malware. Broader system compromise depends on the service account’s privileges and access.

Mitigation

If you can’t patch tonight, do this now.

Until patched, restrict public access to PHP-CGI endpoints or disable the vulnerable CGI integration where feasible. Consider migrating to an alternative PHP integration not exposed to this CGI argument-injection mechanism. Run web services with least privilege and disable unnecessary services to limit post-exploitation impact. For F5 Advanced WAF or BIG-IP ASM deployments, update attack signatures and enable signature 200103548, “PHP-CGI Shell Code Injection.” These controls do not replace patching.

Remediation

Patch, then assume compromise.

The original fixes are PHP 8.1.29, 8.2.20, and 8.3.8 for their respective branches. Upgrade to a current supported, vendor-patched PHP release rather than stopping at these minimum versions: CVE-2024-8926 subsequently bypassed the original fixes under certain nonstandard Windows code-page configurations. That bypass was addressed in PHP 8.1.30, 8.2.24, and 8.3.12. Migrate unsupported branches to a supported release or obtain a maintained vendor backport. For Echo php-7.2 packages, install version 7.2.34+e1 or later.

Public exploits

Exploits

38 exploits validated by Mallory's analysis. Hidden: 35 rejected by Mallory's analysis as not a working exploit; 1 not yet analyzed, such as Exploit-DB entries. Showing the 25 most capable, earliest first.

View more in app →
php_cgi_arg_injection_rce_cve_2024_4577.rbThis repository contains a single Metasploit module implementing a weaponized exploit for CVE-2024-4577, a PHP CGI argument injection vulnerability on Windows. The exploit targets PHP installations running in CGI mode, particularly those with locales that use Unicode best-fit character conversion (e.g., Chinese or Japanese). The module is designed to work against default XAMPP installations, targeting the /php-cgi/php-cgi.exe endpoint, but can be configured for other PHP CGI endpoints. The exploit works by sending a specially crafted POST request with encoded arguments that bypass PHP's argument parsing, allowing arbitrary PHP code or system commands to be executed. The module supports both PHP and Windows command payloads, including Meterpreter shells. The code is structured as a standard Metasploit exploit module, with options for target URI and payload selection. The repository is a single Ruby file, fully integrated into the Metasploit framework, and is operational and weaponized.Disclosed Jun 6, 2024 · Framework: metasploit · ruby · networkWeaponizedCVE-2024-4577-RCE-EXPThis repository contains a Python exploit script (xampp.py) and a README for CVE-2024-4577, a remote code execution vulnerability affecting PHP CGI on Windows systems, specifically those configured in Simplified Chinese, Traditional Chinese, or Japanese. The exploit is designed for bulk exploitation: it reads a list of target URLs from 'xamp.txt', then uses multiple threads (default 1000) to send specially crafted POST requests to common PHP CGI endpoints (/cgi-bin/php-cgi.exe and /php-cgi/php-cgi.exe). The payloads attempt to enable 'allow_url_include' and 'auto_prepend_file' via query parameters, then inject PHP code to test for code execution. If successful, the script writes the vulnerable URLs and responses to output files. The exploit demonstrates code execution by issuing a ping to an attacker-controlled DNS domain (win.kyeezvd0.dnslog.pw), which can be monitored for callbacks. The README provides usage instructions and highlights the need to adjust the command, input file, and thread count as needed. The repository is operational and suitable for real-world exploitation, not just proof-of-concept.WanLiChangChengWanLiChang · Disclosed Jun 7, 2024 · python · networkOperationalCVE-2024-4577This repository is a Python-based exploit for CVE-2024-4577, a PHP CGI argument injection vulnerability affecting certain Windows configurations (notably XAMPP and similar stacks). The exploit consists of three files: a README.md with usage and background, a requirements.txt listing Python dependencies, and the main exploit script exploit.py. The exploit.py script allows users to scan single or multiple URLs for the vulnerability, and if a target is found to be vulnerable, it provides an interactive shell for executing arbitrary system commands via the injected PHP payload. The exploit works by sending a specially crafted POST request to the /php-cgi/php-cgi.exe endpoint, abusing Unicode best-fit conversion to bypass argument restrictions and inject PHP settings and code. The tool is multi-threaded for bulk scanning and can save lists of vulnerable URLs. The exploit is operational and provides real command execution on vulnerable targets, but is not part of a larger exploitation framework.Chocapikk · Disclosed Jun 9, 2024 · python · networkOperationalcve-2024-4577This repository contains a Python proof-of-concept exploit for CVE-2024-4577, a PHP-CGI argument injection vulnerability. The main file, cve-2024-4577.py, allows a user to test a remote server for the vulnerability and, if present, execute arbitrary PHP code via a crafted POST request to the /php-cgi/php-cgi.exe endpoint. The script accepts a target URL and an optional custom PHP payload, defaulting to phpinfo() for detection. If a custom payload is provided, it can be used to execute system commands (e.g., whoami) on the target. The exploit is operational and can be used for both detection and exploitation. The repository also includes a brief readme.md with usage instructions. No hardcoded IPs or domains are present; the endpoint is constructed based on user input.aavamin · Disclosed Jun 12, 2024 · python · networkOperationalCVE-2024-4577This repository contains a Python exploit script (CVE-2024-4577.py) targeting the argument injection vulnerability in PHP (CVE-2024-4577). The script allows the user to scan a single target or a list of targets (provided in 'domains.txt') for the vulnerability. It works by sending a specially crafted POST request to the '/index.php' endpoint of the target, attempting to inject and execute arbitrary PHP code via the 'auto_prepend_file' argument. The exploit checks for successful code execution by looking for a specific string in the response. The repository includes a README with usage instructions and a requirements.txt for dependencies. The main exploit file is self-contained, multithreaded for scanning multiple domains, and operational, as it allows arbitrary command execution if the target is vulnerable.gotr00t0day · Disclosed Jun 15, 2024 · python · networkOperationalCVE-2024-4577-RCEThis repository provides a Python-based exploit and scanner for CVE-2024-4577, a PHP CGI Argument Injection vulnerability affecting PHP versions 8.1.* before 8.1.29, 8.2.* before 8.2.20, and 8.3.* before 8.3.8 on Windows systems. The main file, CVE-2024-4577.py, can scan a target for vulnerability and exploit it by sending a specially crafted HTTP POST request to a PHP CGI endpoint (such as /php-cgi/php-cgi.exe), injecting arbitrary PHP code provided by the attacker. The exploit supports custom PHP payloads, including reverse shells. The README provides detailed usage instructions, example payloads, and references. The repository is operational and can be used to achieve remote code execution on vulnerable servers. The only code file is in Python, and the exploit is not part of a larger framework.l0n3m4n · Disclosed Jul 6, 2024 · python · networkOperationalCVE-2024-4577-RCE-ATTACKThis repository provides a proof-of-concept (PoC) and operational exploit for CVE-2024-4577, a critical remote code execution vulnerability affecting PHP (versions 8.1.x < 8.1.29, 8.2.x < 8.2.20, 8.3.x < 8.3.8) when run as CGI on Windows (commonly with Apache, e.g., XAMPP). The main exploit script (CVEs-2024-4577.py) is written in Python and can both scan for and exploit the vulnerability. It works by sending specially crafted HTTP POST requests to vulnerable endpoints (such as /php-cgi/php-cgi.exe, /index.php, etc.), injecting PHP options via command-line arguments to enable remote code execution. If exploitation is successful, the script can upload and execute a PHP payload (rev_shell.php), which is a reverse shell that connects back to the attacker's machine, granting remote command execution. The repository includes a requirements.txt for dependencies, a sample reverse shell payload (rev_shell.php), and a detailed README.md with usage instructions, references, and detection tips. The exploit is operational and can be used for both vulnerability scanning and full exploitation, depending on the provided arguments.bibo318 · Disclosed Jul 11, 2024 · python · php · networkOperationalCVE-2024-4577This repository provides a working exploit for CVE-2024-4577, a critical remote code execution vulnerability affecting PHP on Windows (versions 8.3 < 8.3.8, 8.2 < 8.2.20, 8.1 < 8.1.29) when PHP-CGI is exposed. The repository contains two main Python scripts: - CVE_2024_4577_POC.py: A proof-of-concept scanner that reads a list of URLs from 'urls.txt' and tests each for the vulnerability by sending a crafted POST request to the '/php-cgi/php-cgi.exe' endpoint. It checks for code execution by looking for a specific string in the response. - CVE_2024_4577_exp.py: An exploit script that allows the user to specify an arbitrary system command to execute on the vulnerable server, again via a POST request to the same endpoint. The PHP payload is injected via the request body and executed by the server if vulnerable. The README provides brief usage instructions. The exploit leverages the ability to pass PHP parameters via the query string to the CGI binary, bypassing restrictions and enabling arbitrary code execution. The attack vector is network-based, requiring only HTTP access to the vulnerable endpoint. The scripts are operational and allow for both detection and exploitation of the vulnerability.Jcccccx · Disclosed Jul 31, 2024 · python · networkOperationalCVE-2024-4577-RCEThis repository provides a Python-based exploit and scanner for CVE-2024-4577, a PHP CGI Argument Injection vulnerability affecting PHP versions 8.1.x before 8.1.29, 8.2.x before 8.2.20, and 8.3.x before 8.3.8 on Windows systems running Apache with PHP-CGI. The main script, CVE-2024-4577.py, can scan a list of target URLs for the vulnerability and, if found, exploit it by sending arbitrary PHP code (such as a reverse shell) via a specially crafted HTTP POST request. The exploit leverages a flaw in how PHP-CGI parses command-line arguments, allowing attackers to inject PHP options and execute code. The README provides detailed usage instructions, including how to prepare a PHP payload for remote code execution and how to use the tool for both scanning and exploitation. The repository is operational, requiring the attacker to supply a PHP payload file, and is capable of establishing a reverse shell to a specified host and port. The exploit targets network-accessible HTTP endpoints, particularly /php-cgi/php-cgi.exe, and is especially effective against default XAMPP installations on Windows.gh-ost00 · Disclosed Aug 20, 2024 · python · networkOperationalCVE-2024-4577This repository contains a Python-based scanner and exploit tool for CVE-2024-4577, a critical PHP CGI argument injection vulnerability that can lead to remote code execution (RCE). The main script, CVE-2024-4577.py, provides both scanning and exploitation capabilities. It reads a list of target URLs from a file, checks each for vulnerability by sending a crafted POST request, and, if vulnerable, can exploit the target by sending arbitrary PHP code (provided by the user in a separate file) to be executed on the server. Vulnerable targets are logged to 'victim.txt'. The tool is command-line driven, supports both scanning and exploitation modes, and uses the 'requests', 'colorama', and 'rich' Python libraries for HTTP requests and output formatting. The README provides detailed usage instructions, search queries for finding potential targets, and a disclaimer. The exploit targets PHP CGI installations on both Linux and Windows platforms, specifically versions 8.1, 8.2, and 8.3.longhoangth18 · Disclosed Oct 14, 2024 · python · networkOperationalCVE-2024-4577-RCE-PoCThis repository contains a Python exploit script (CVE-2024-4577.py) and a README.md for CVE-2024-4577, a critical argument injection vulnerability in PHP running on Windows (notably in XAMPP environments). The exploit targets php-cgi.exe via HTTP(S) requests, leveraging encoding conversion issues to bypass previous mitigations and achieve unauthenticated remote code execution (RCE). The script supports both direct and blind RCE, with options to save output and test various payload delivery methods (php://input, data://, php://filter). It can also set up a local listener to detect outbound requests from the target in blind RCE scenarios. The README provides detailed usage instructions, exploitation conditions, and background on the vulnerability. The main fingerprintable endpoint is '/php-cgi/php-cgi.exe', which is used for the attack. The exploit is operational, allowing arbitrary PHP code execution on vulnerable targets.BTtea · Disclosed Nov 6, 2024 · python · markdown · networkOperationalCVE-2024-4577This repository provides operational exploit scripts for CVE-2024-4577, a remote code execution vulnerability affecting PHP 5 and 7. The repository contains two main Python scripts: 'cek.py' and 'cuk.py'. - 'cek.py' allows an attacker to execute arbitrary system commands on a vulnerable PHP server by exploiting the allow_url_include and auto_prepend_file directives via a crafted POST request. It supports listing directories, reading files, and finding writable directories on the target server. - 'cuk.py' is used to upload a remote PHP shell ('ini.php') from a GitHub repository to the target server's /tmp directory, enabling persistent access. Both scripts require the attacker to specify a target URL (typically an index.php endpoint) and leverage a POST request with a malicious PHP payload. The exploit is network-based and targets publicly accessible PHP endpoints. The repository is structured with a README providing usage instructions and two Python scripts implementing the exploit logic. No detection scripts or fake exploits are present; the code is functional and provides real RCE capabilities.Dejavu666 · Disclosed Jan 8, 2025 · python · networkOperationalCVE-2024-4577This repository contains a Python exploit script (CVE-2024-4577.py) targeting the argument injection vulnerability in PHP (CVE-2024-4577). The script allows the user to scan a single target or a list of targets (provided in domains.txt) for the vulnerability by sending a specially crafted POST request to the /index.php endpoint of each target. The payload, supplied via the -c/--command argument, is arbitrary PHP code that will be executed on the server if the target is vulnerable. The script checks for successful code execution by looking for a specific marker ('gotr00t') in the response. The repository also includes a README.md with usage instructions and a requirements.txt listing the Python dependencies (requests, colorama). The exploit is operational and can be used to achieve remote code execution on vulnerable PHP installations accessible over the network.Didarul342 · Disclosed Feb 14, 2025 · python · networkOperationalphp-cgi-InjectorThis repository is a comprehensive exploit toolkit targeting PHP-CGI parameter injection vulnerabilities, specifically CVE-2024-4577 and CVE-2024-8926. The main exploit logic resides in 'exploit.py', which provides a command-line interface for automated vulnerability scanning and exploitation. The tool supports multiple attack modes, including remote command execution, arbitrary PHP code execution, file upload, and file download. It allows users to specify target URLs and CGI endpoints, select or customize payloads, and optionally route traffic through Tor for anonymity. A notable feature is the modular bypass system: the 'bypass_manager.py' and the 'bypass_modules/' directory provide various tamper scripts to obfuscate payloads and evade WAFs or security controls. These modules can modify payloads, CGI paths, HTTP headers (e.g., randomizing User-Agent, adding X-Forwarded-For), and more. The tool also supports forced exploitation mode, detailed logging, and animated terminal output for usability. The repository is structured as follows: - 'exploit.py': Main exploit script and entry point. - 'bypass_manager.py' and 'bypass_modules/': Framework for WAF bypass and payload tampering. - 'tor_helper.py': Utilities for routing requests through Tor. - 'requirements.txt': Python dependencies. - 'README.md': Detailed usage instructions, attack modes, and legal disclaimer. The exploit is operational, providing real attack capabilities against vulnerable PHP-CGI deployments. It is not a detection-only script, but a full-featured exploitation tool with customizable payloads and bypass techniques. The main attack vector is network-based, targeting web servers exposing PHP-CGI endpoints. The tool is suitable for red team operations, CTFs, and security research.Night-have-dreams · Disclosed Mar 15, 2025 · python · networkOperationalCVE-2024-4577-ExploitThis repository provides a Python-based exploit for CVE-2024-4577, a PHP CGI argument injection vulnerability that enables remote code execution (RCE) on vulnerable PHP installations running in CGI mode on Windows. The main exploit logic is contained in 'main.py', which accepts a list of target URLs (either as a comma-separated string or from a file) and a command to execute. The script sends a specially crafted POST request to each target, exploiting the argument injection flaw to execute arbitrary system commands. The exploit supports parallel execution across multiple CPU cores for efficiency and logs successfully exploited targets to a file. The only code file is 'main.py', with dependencies listed in 'requirements.txt'. The attack vector is network-based, targeting HTTP endpoints running vulnerable PHP CGI. The exploit is operational, requiring the attacker to specify the command to execute, and provides full RCE if successful. No hardcoded IPs or domains are present; the targets are user-supplied.Gill-Singh-A · Disclosed Apr 18, 2025 · python · networkOperationalCVE-2024-4577This repository contains a single Python exploit script (CVE-2024-4577.py) targeting the PHP CGI Argument Injection Remote Code Execution vulnerability (CVE-2024-4577). The exploit is operational and allows an attacker to execute arbitrary system commands on a vulnerable server by abusing improper argument handling in PHP CGI. The script supports both single-target and multi-target modes, using HTTP POST requests to various common PHP CGI endpoints (e.g., /php-cgi/php-cgi.exe, /cgi-bin/php-cgi.exe, etc.). It injects PHP configuration arguments via specially crafted URLs and delivers a PHP payload that executes attacker-supplied commands. The exploit is multi-threaded for mass scanning and includes output formatting and progress indication. The README is minimal and only references the exploit file. No detection-only or fake code is present; the script is a real exploit for RCE. The main attack vector is network-based, requiring HTTP(S) access to a vulnerable PHP CGI endpoint.ibrahmsql · Disclosed Jun 15, 2025 · python · networkOperationalCVE-2024-4577This repository provides tools for detecting and exploiting CVE-2024-4577, a critical PHP CGI argument injection vulnerability affecting Windows systems running PHP in CGI mode. The repository contains two main Python scripts: 1. CVE-2024-4577.py: A scanner that tests whether a given URL (or list of URLs) is vulnerable by sending a PHP payload and checking for a known hash in the response. It targets common CGI endpoints such as /php-cgi/php-cgi.exe, /index.php, /test.php, and /test.hello, using specially crafted query strings to trigger the vulnerability. 2. exploit.py: An exploit tool that allows the user to execute arbitrary system commands on a vulnerable server. It sends a PHP payload via HTTP POST to the vulnerable endpoint, which executes the specified command on the server and returns the output. The repository also includes a requirements.txt for dependencies and a README.md with detailed usage instructions, example commands, and background on the vulnerability. The exploit is operational, providing real command execution on vulnerable targets, and is not part of a larger framework. The main attack vector is network-based, targeting HTTP(S) endpoints exposed by PHP CGI on Windows. The endpoints tested are fingerprintable and can be used to identify potentially vulnerable systems.r0otk3r · Disclosed Jul 7, 2025 · python · networkOperationalctf-cve-2024-4577This repository is a comprehensive CTF challenge environment demonstrating exploitation of CVE-2024-4577, a critical PHP CGI argument injection vulnerability. The structure includes: - **attack-scripts/**: Contains Python and Bash scripts (`exploit.py`, `exploit.sh`, `container-escape.py`) to automate exploitation, file reading, command execution, container escape, and AWS pivoting. - **vulnerable-app/**: A minimal PHP web application (`index.php`, `info.php`) and Dockerfile, intentionally misconfigured to run PHP in CGI mode and allow argument injection. The flag file (`/flag.txt`) is the main target. - **kubernetes/**: Manifests for deploying the vulnerable app in a Kubernetes cluster, with privileged containers, host mounts (`/host`, `/dev`, `/proc`, `/sys`), and Docker socket exposure for advanced exploitation and container escape scenarios. AWS credentials are provided via ConfigMap for cloud pivoting. - **docker/**: Docker Compose and frontend files for local or CTF deployment. - **Documentation**: Multiple guides and walkthroughs detail exploitation steps, deployment, and security considerations. **Exploit Capabilities:** - Exploits CVE-2024-4577 via HTTP requests with crafted PATH_INFO to inject PHP command-line arguments, enabling arbitrary file read (e.g., `/flag.txt`), command execution, and code execution. - Advanced scripts attempt container escape by accessing host filesystems and Docker socket, and pivot to AWS using provided credentials. - Can establish a reverse shell if the attacker provides a listener IP and port. **Endpoints:** - Main HTTP endpoint (default: `http://localhost:8080` or via Kubernetes LoadBalancer/Ingress) - File targets: `/flag.txt`, `/etc/passwd`, `/host`, `/var/run/docker.sock`, and others - AWS resources (via fake credentials in the CTF) **Purpose:** The repository is designed for educational and CTF use, providing a realistic, multi-stage exploitation chain: initial web exploit, container escape, and cloud pivoting. It demonstrates the risks of insecure PHP CGI configurations, container privilege escalation, and cloud credential exposure. The environment is highly configurable for both local and internet-facing deployments, with extensive documentation and security warnings.a1ex-var1amov · Disclosed Aug 25, 2025 · python · bash · networkOperationalCVE-2024-4577-newThe supplied repository contains four files: two standalone Python entry points, a README, and pinned Python dependencies. CVE-2024-4577.py is an active exploitation-based scanner supporting a single URL, a target-list file, and an optional custom path. It sequentially POSTs a PHP MD5 probe to four default routes with encoded soft-hyphen argument injection. The query attempts to change CGI redirect settings, enable URL inclusion, and prepend php://input. Detection depends solely on finding the hardcoded string 3f2ba4ab3b260f4c2dc61a6fac7c3e8a in the response; the script does not compute its expected value dynamically. Custom paths are tested as supplied without automatically adding the injection query. exploit.py accepts a required command and either a single target or a target-list file, skips blank/comment lines in list mode, and submits a PHP system() payload. However, it always targets the bare /php-cgi/php-cgi.exe endpoint without the necessary argument-injection parameters. Consequently, it is an incomplete CVE-2024-4577 command exploit under normal vulnerable-server conditions. It also labels any HTTP 200 response as successful, and directly interpolates commands into a PHP double-quoted string without escaping, which can break payload syntax or alter command contents. The README's manual curl example has the same missing-query limitation. Both scripts use requests with a 10-second timeout and disabled TLS certificate verification. Target addresses and input-file paths are operator supplied; there are no hardcoded operational domains, IP addresses, callbacks, or exfiltration endpoints. requirements.txt pins requests and its dependencies. README.md provides usage examples, screenshots, social links, and a license claim, but the referenced LICENSE file is absent from the supplied inventory. No exploit framework is present. The supplied file sizes total 9,086 bytes; the repository URL, git reference, archive path, and actual archive size were not provided, so repository metadata uses empty or unknown placeholders.lockbitblack-ai · Disclosed Feb 11, 2026 · python · web · networkOperationalPHP_8.1.x_ExploitThis repository is a standalone Bash-based exploit toolkit centered on PHP8.1.x_Exploit.sh, which orchestrates detection and exploitation of four PHP CVEs using modular helper scripts under modules/. The main script normalizes target URLs, fingerprints OS and PHP version via HTTP headers and a phpinfo.php probe, crawls homepage links/forms, enumerates likely CGI and upload endpoints, and then invokes per-CVE modules. The repository structure is simple: one main driver script, four CVE-specific Bash modules, one shared utils module, and a README. Primary exploit capability is CVE-2024-4577 against Windows PHP-CGI. The module tests two query-string injection variants, including a soft-hyphen bypass, by setting allow_url_include and auto_prepend_file=php://input and POSTing PHP code. On success it either demonstrates command execution with whoami or launches a PowerShell reverse shell to an operator-supplied IP/port. This is a real exploitation path providing RCE. A second remote capability targets CVE-2025-14177 by generating malformed JPEG files and POSTing them to discovered upload endpoints using multipart form field file. Detection looks for a marker string or long printable output in responses; exploitation repeats the upload 10 times and stores raw responses and extracted strings under a reports subdirectory, aiming to capture leaked heap data. The remaining two modules, CVE-2025-14180 and CVE-2025-14178, are effectively local/RCE-follow-on checks and crash payloads rather than pure remote exploits. Both use php -r locally when PHP CLI is present; the exploit functions POST PHP snippets to a target URL if remote code execution is already available. CVE-2025-14180 attempts to trigger a PDO PostgreSQL null dereference using DSN pgsql:host=localhost;dbname=postgres with emulated prepares enabled. CVE-2025-14178 attempts an array_merge integer overflow by allocating extremely large arrays. Both are described as potentially crashing or destabilizing the PHP process. Overall, this is an operational multi-CVE web exploitation script, not just a detector. It supports single-target and batch modes, optional GNU parallel execution, report logging, and automatic exploitation unless scan-only is specified. Fingerprintable targets include common CGI paths, common upload paths, phpinfo.php probing, php://input injection, localhost PostgreSQL access, and attacker-defined reverse-shell callback endpoints.gl1tch0x1 · Disclosed Apr 29, 2026 · bash · markdown · web · network · localOperationalcve-2024-4577-phpcgi_rce_reproductionThis repository is a small standalone Python exploit for CVE-2024-4577, a Windows-specific PHP CGI argument injection vulnerability known as 'WorstFit'. The repository contains only two files: a README with vulnerability background, affected versions, setup notes, and usage examples; and a single executable script, exploit.py, which is the main entry point. The exploit works by appending crafted query-string arguments to a user-supplied php-cgi.exe URL. It uses the soft-hyphen encoded sequence %AD to bypass the legacy CGI hyphen filter via Windows Best-Fit character conversion, causing PHP CGI to interpret the query string as command-line options. Specifically, it injects directives equivalent to '-d allow_url_include=1' and '-d auto_prepend_file=php://input'. The script then sends attacker-controlled PHP code in the HTTP request body. In command mode, the body is '<?php system('<cmd>'); ?>', yielding arbitrary OS command execution. In phpinfo mode, it sends '<?php phpinfo(); ?>' to confirm code execution and disclose environment details. Code structure is simple: build_payload() constructs the malicious query string and PHP body; exploit() sends either a GET request for phpinfo mode or a POST request for command execution, with optional proxy support and disabled TLS verification; main() parses CLI arguments and enforces that either --cmd or --phpinfo is supplied. The exploit does not include target discovery, brute forcing, persistence, or post-exploitation automation. It is therefore best classified as an operational standalone RCE exploit rather than a framework module or detection-only script.razureink · Disclosed Jul 23, 2026 · markdown · python · web · networkOperationalcve-2024-4577-labThis 13-file repository is a self-contained Docker/CTF laboratory for PHP-CGI argument injection. Its main operational entry point, exploit.sh, uses curl to POST a PHP payload to /index.php while injecting php-cgi -d options through a query string. The injected options set allow_url_include=1 and auto_prepend_file=php://input, causing PHP to execute the request body before the normal application page. The payload reports context and executes a supplied shell command; its default reads the container flag. Dockerfile builds intentionally unpatched PHP 5.4.1 with CGI support on Debian Bullseye, installs it as /usr/lib/cgi-bin/php-cgi, and configures Apache mod_actions to route .php files to that binary. config/apache-vhost.conf is central to the vulnerability: CGI query-string words become argv, and old php-cgi parses those attacker-controlled words as command-line options. config/php.ini disables cgi.force_redirect and leaves allow_url_include disabled by default so the exploit must enable it at runtime. Dockerfile.vulhub provides a fallback image, docker-compose.yml publishes container port 80 as localhost:8080, start.sh launches Apache, app/index.php is a benign status page, and poc.http provides raw Linux and Windows-form requests. The repository accurately distinguishes its runnable Linux target from real CVE-2024-4577. The Linux container uses literal '-' option injection against pre-2012-patch PHP 5.4.1, therefore demonstrating CVE-2012-1823's underlying CGI argument-injection/RCE primitive. poc.http also documents the %AD soft-hyphen form used by CVE-2024-4577 on affected Windows PHP-CGI systems, where Windows best-fit decoding converts %AD to '-' after the original hardening check.khwajasaad267-coder · Disclosed Aug 25, 2026 · bash · php · webOperationaledu-reconedu-recon is a Python 3.10+ reconnaissance, vulnerability-triage, and active assessment orchestrator aimed at education-sector engagements. It has 38 files, primarily Python modules, with `recon.py` as the CLI entry point and `run.sh` as a one-command launcher. Its stdlib web UI (`edurecon/webui.py`) exposes scan submission, live logs, findings, report/reproduction generation, and dump functions; it defaults to localhost but can be bound to all interfaces without authentication. The engine expands CIDRs through nmap ping sweeps and domains through subdomain enumeration, then concurrently runs Shodan enrichment, nmap service/version scanning, dirsearch web discovery, exposure checks, secret/API-key scanning, Moodle auditing, reflected-XSS and SQLi checks, Hydra credential attacks, and WordPress wp2shell processing. Full intensity actively invokes sqlmap and Hydra; recon mode lists those candidates instead. It uses external repositories installed by `setup` for PHP-CGI, React2Shell, WordPress, and directory-scanning capabilities. Built-in CVE probes target PHPUnit eval-stdin RCE (CVE-2017-9841), Apache traversal/LFI (CVE-2021-41773), Struts S2-045 (CVE-2017-5638), Confluence OGNL RCE (CVE-2022-26134), Drupalgeddon2 (CVE-2018-7600), Next.js middleware bypass (CVE-2025-29927), PHP-CGI (CVE-2024-4577/CVE-2024-8926), and React Server Components RCE (CVE-2025-55182). Several internal checks use benign confirmation markers/arithmetic oracles, but the overall repository is an active exploitation platform: full mode can perform credential guessing and injection, external wp2shell can pursue SQLi-to-shell, and React2Shell can be configured away from safe-check mode to run a command. Notable collection functionality includes downloading exposed files and backups, scanning their contents for cloud/VCS/payment/API credentials, and `gitdump.py`, which mirrors an exposed `.git` directory, retrieves reachable loose Git objects, and reconstructs source files. Artifacts, reports, findings, and potentially sensitive material are stored under `runs/<run-id>/`; report code serializes evidence and can therefore retain sensitive values. ScopeGuard supports target/CIDR and subdomain allowlisting, but supplied documentation/configuration describe scope enforcement as disabled by default, materially increasing operator-supplied target risk.yeee3642 · Disclosed Aug 26, 2026 · python · bash · network · webOperationalphp-exploits-CVE-2026-109618-CVE-2026-109619-CVE-2024-4577The repository contains 11 files: three CVE directories, each with a Python poc.py, an extended exploit.py, and Portuguese technical documentation, plus a root README and an Unlicense. All six scripts have CLI entry points. They use requests and urllib3, disable TLS certificate verification in visible sessions, and support optional proxies. No established exploit framework is identified. The advertised capabilities exceed detection, so the repository is classified as operational exploit tooling, but this does not establish that every implementation works. The CVE-2024-4577 code follows the recognizable Windows PHP-CGI argument-injection mechanism: a soft-hyphen query injects PHP configuration flags, and auto_prepend_file=php://input makes the POST body executable PHP. Visible code probes three CGI paths with literal-hyphen and soft-hyphen requests; poc.py lists nine discovery candidates. Command execution, system information, file reads, and an HTTP pseudo-shell are exposed through dispatch code. The extended version also dispatches reverse shells, file transfers, webshell placement, and internal-network probing, although their implementations are largely truncated. Its locale probe demonstrates differential payload behavior, not identification of an exact Windows code page. The two 2026 directories claim a 'php2shell' family involving OOM-induced HashTable corruption and an upload-progress use-after-free. Their documentation describes large POST bodies, heap grooming, concurrent multipart uploads, shared PHPSESSID values, and PHP_SESSION_UPLOAD_PROGRESS spray content. However, the supplied excerpts omit the core memory-exploitation implementations. PHP code placed in form fields or heap-spray data is not inherently executable. HTTP 5xx responses, timing anomalies, or reflected markers would not independently prove memory corruption, disclosure, or RCE. CVE validity, affected-version ranges, patch releases, active exploitation, and linked advisories for these two identifiers remain unverified repository assertions. Visible post-exploitation helpers in both 2026 exploit files perform enumeration rather than actual privilege escalation. They optionally attempt cron-based reverse-shell persistence, create an SSH directory, and inspect Apache webroot configuration. Both reverse-shell dispatch branches set the JSON rce field to true without a visible success check, making that output unreliable as proof of compromise. In CVE-2026-109619/poc.py, --exploit routes to the same race demonstration method as --race-demo, rather than a separately visible RCE implementation. Fingerprintable activity includes CGI requests containing %ADd, allow_url_include and auto_prepend_file; LOCALE_OK_ probe responses; large parameter-heavy POSTs; and bursts of multipart uploads and reads sharing PHPSESSID. Targets, callback addresses, proxy settings, and output paths are operator-controlled. alvo.com, private IP addresses, and localhost proxy URLs are examples, not established attacker infrastructure. Advisory and author URLs are documentation links, with no visible runtime calls to them. All six source files are truncated, preventing full validation of payloads and endpoint coverage. The root README's suggested Linux php:8.3.4-apache container does not reproduce the Windows Best-Fit prerequisite of CVE-2024-4577, and its referenced docker configuration file is absent from the inventory. No affirmative fake-exploit classification is warranted from the excerpts alone, but the 2026 RCE claims need substantial verification. No original repository URL, git reference, or archive size was supplied; the empty metadata strings and size_bytes=0 represent unavailable metadata. Listed file sizes total 150,899 bytes, which is not an archive-size measurement.Ghost-Bx1 · Disclosed Oct 23, 2026 · python · web · networkOperationalPOC-CVE-2026-109618-CVE-2026-109619-CVE-2024-4577Standalone collection containing six Python 3 command-line programs, four Markdown documents and an Unlicense file. Each CVE directory contains poc.py, exploit.py and a Portuguese technical README. Python orchestrates HTTP requests and embeds PHP payloads. Visible dependencies include requests and urllib3; documentation additionally requests requests-toolbelt. All six source excerpts contain omitted sections, so this is a static, partial-source assessment rather than execution-based validation. Repository URL, git reference and archive size were not supplied; empty metadata and size 0 represent unknown values. The CVE-2024-4577 implementation contains a recognizable exploitation primitive: POST to a PHP-CGI endpoint with soft-hyphen argument injection enabling allow_url_include and auto_prepend_file=php://input. Visible code supports arbitrary PHP submission, command execution and endpoint discovery, and embeds a command/upload webshell. CLI dispatch advertises reverse shells, file transfers, system information and internal scanning, but the corresponding implementations are mostly omitted. Detection is unreliable: verificar accepts any nonempty HTTP 200 response without literal PHP source as an alternative to its marker check, allowing false positives. Discovery likewise treats many non-404/non-400 responses as endpoint hits without establishing exploitability. The CVE-2026-109618 files describe heap grooming followed by oversized parameter submissions to provoke OOM, response-based RCE validation and memory-leak searches. Visible code establishes large request construction, not a controlled memory-corruption-to-execution chain. The CVE-2026-109619 files describe concurrent multipart uploads and reads sharing PHPSESSID, with version-dependent spray sizes. Its visible spray builder merely wraps Base64 PHP or filler in serialized-looking text; no native forged structures, addresses or control-flow redirection are shown. In poc.py, --exploit invokes the same race demonstration as --race-demo, changing the iteration count rather than selecting a demonstrated RCE primitive. HTTP 5xx responses, timing anomalies or reflected markers do not prove UAF, memory disclosure or execution. The 2026 CVE records, release versions, exploitation reports and advisory links remain repository claims, not independently verified facts. Requests disable TLS certificate validation and support optional proxies. Distinctive observables include %ADd, allow_url_include, auto_prepend_file, direct PHP-CGI requests, PHP_SESSION_UPLOAD_PROGRESS, shared PHPSESSID values, SPRAY_<size> strings, and the markers OOM_RCE_CONFIRMED_CVE2026109618 and UAF_RCE_OK_CVE2026109619. No fixed external exfiltration host or malicious operator-side action is visible. Documentation URLs are references, not demonstrated runtime callbacks. Large-body and concurrent-request modes can disrupt services even when exploitation fails. The README's Linux php:8.3.4-apache Docker example does not reproduce the Windows-specific Best-Fit prerequisite for CVE-2024-4577. Overall, the repository contains a substantive CGI exploit alongside unproven memory-corruption exploit claims; the collection is not classified as wholly fake, but its advertised capabilities should not be accepted at face value.Ghost-Bx1 · Disclosed Oct 23, 2026 · python · php · webOperational
Exposure surface

Affected products & vendors

Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.

VendorProductType
Fedora ProjectFedoraoperating_system
PHP GroupPhpapplication
FAQ

Frequently asked questions

Is CVE-2024-4577 exploited in the wild?
Yes. As of Nov 13, 2026, Mallory has reports of exploitation in the wild, and CISA added it to KEV on Jun 12, 2024.
Is there a public exploit for CVE-2024-4577?
Yes. Mallory tracks 38 public exploits or proofs of concept for CVE-2024-4577, the earliest disclosed Jun 6, 2024.
Is there a patch for CVE-2024-4577?
The original fixes are PHP 8.1.29, 8.2.20, and 8.3.8 for their respective branches. Upgrade to a current supported, vendor-patched PHP release rather than stopping at these minimum versions: CVE-2024-8926 subsequently bypassed the original fixes under certain nonstandard Windows code-page configurations.
What products are affected by CVE-2024-4577?
PHP Group Php and Fedora Project Fedora.
How severe is CVE-2024-4577?
CVE-2024-4577 has a CVSS v3.1 score of 9.8 (Critical) and an EPSS score of 100.0% (100th percentile) as of Nov 13, 2026.
Activity feed

Recent activity

106 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.

View more in app →
+74 more in the timeline. Sign up to see them
Do you run this?See affected versions in your environment and whether adversaries are exploiting this vulnerability against assets like yours.
Check your exposure
Mallory platform

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: which of your assets are affected, which adversaries are exploiting it right now, and which detections to deploy tonight.
Exposure mappingQuery your assets running an affected version, and investigate the blast radius.
Threat actor evidence3Every report linking this CVE to a named adversary.
Associated malware42Malware families linked to this CVE, with evidence and IOCs.
Detection signatures5YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Vendor-by-vendor mappingCross-references every affected SKU, including bundled OEM variants.
Social activity19Community discussion across Reddit, Mastodon, and other social sources.