CVE-2026-48495 in Typebot
Summary
by MITRE • 08/11/2026
TypeBot is a chatbot builder tool. Prior to version 3.17.0, the Google Sheets OAuth callback decodes a base64-encoded JSON `state` parameter and trusts the embedded `workspaceId`, `typebotId`, `blockId`, and `redirectUrl` without cryptographic integrity protection or authorization checks. The callback route is authenticated, but it does not verify that the authenticated user has write access to the target workspace or Typebot before creating credentials in the workspace or updating Typebot groups. An authenticated user who can obtain a valid Google OAuth `code` can alter the `state` value to create Google Sheets credentials in another workspace and, if target IDs are known, attach those credentials to a block in another Typebot. Version 3.17.0 patches the issue.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 08/11/2026
The vulnerability in TypeBot affects versions prior to 3.17.0 and stems from a critical flaw in the Google Sheets OAuth callback implementation that violates fundamental security principles of authentication and authorization. This vulnerability represents a classic case of insufficient input validation and trust model abuse where the system blindly accepts and processes data from external sources without proper verification mechanisms.
The technical flaw manifests in how the OAuth callback handles the base64-encoded JSON state parameter which contains critical identifiers including workspaceId, typebotId, blockId, and redirectUrl. The system performs no cryptographic integrity checks or authorization validation before processing these values, creating a path for privilege escalation attacks. This design pattern aligns with CWE-295 which addresses improper certificate validation and CWE-347 which covers insufficient cryptographic padding. The absence of proper authorization checks means that any authenticated user who can manipulate the state parameter can effectively impersonate access to other users' workspaces and typebots.
The operational impact of this vulnerability is significant as it allows an attacker to perform unauthorized modifications across different workspaces and typebots within the same system. An authenticated user who has obtained a valid Google OAuth code can modify the state parameter to reference target workspace and typebot identifiers, enabling them to create Google Sheets credentials in another user's workspace and attach those credentials to blocks within other typebots. This essentially enables cross-user privilege escalation and data manipulation capabilities that could lead to unauthorized access to sensitive information and potential system compromise.
The vulnerability is particularly concerning because it operates at the intersection of authentication and authorization boundaries where the system assumes that any user with a valid OAuth code can manipulate any target resource they specify in the state parameter. This violates the principle of least privilege and demonstrates a dangerous trust model where external parameters are treated as authoritative without proper validation. The fix implemented in version 3.17.0 addresses these issues by introducing proper cryptographic integrity checks and authorization verification before allowing modifications to workspace credentials or typebot configurations.
From an attack perspective, this vulnerability maps to several ATT&CK techniques including TA0001 Initial Access through credential compromise, TA0002 Execution via privilege escalation, and TA0003 Persistence through unauthorized configuration changes. The attack vector is particularly insidious because it leverages legitimate OAuth flows while exploiting the trust relationship between the application and external services to gain unauthorized access to other users' data and configurations. Organizations should prioritize updating to version 3.17.0 or later to prevent exploitation of this vulnerability, and implement additional monitoring for unauthorized credential creation attempts within their TypeBot environments.