Typebot EUVD-2026-56243

Exploitation requires: (1) a valid authenticated Typebot session - the callback route itself enforces authentication, so unauthenticated abuse is not possible; (2) the ability to complete a Google OAuth authorization-code flow and obtain a valid Google OAuth code, which is achievable by any authenticated Typebot user who has connected or attempts to connect their own Google account; (3) knowledge of the target workspaceId, and optionally typebotId and blockId, to inject credentials into a specific victim resource. The limiting factors are the PR:L authentication requirement (eliminates anonymous attackers) and the need for target resource IDs (limits opportunistic scanning but does not prevent targeted abuse between users who share any context, such as invitations or public bot links that may expose IDs). Non-default configuration is not required - any Typebot instance with Google Sheets integration enabled is vulnerable.

The provided CVSS vector (AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:H/A:N) scores 7.1 and accurately reflects the threat profile: network-reachable, low complexity, requires a valid authenticated session (PR:L), with high integrity impact. The C:L component likely accounts for potential indirect exposure of configuration or credential metadata across workspace boundaries. One meaningful friction point not reflected in CVSS is that successful exploitation against a specific target workspace requires the attacker to know that workspace's internal IDs (workspaceId, typebotId, blockId); this limits opportunistic mass exploitation but does not prevent targeted abuse, especially in multi-tenant cloud deployments where IDs may be discoverable through shared links or invitations. No EPSS data was provided in the input, so exploitation probability cannot be numerically assessed. No CISA KEV listing and no public POC are known at time of analysis, placing this in a monitor-and-patch posture rather than emergency response, but the fix is already available and upgrade is straightforward.

An authenticated Typebot user on a shared or multi-tenant instance initiates a legitimate Google Sheets OAuth flow, intercepts or reconstructs the state parameter, re-encodes it as base64 JSON with the workspaceId and typebotId of a victim workspace they have no access to, and completes the Google OAuth code exchange. The callback processes the tampered state and stores the attacker's Google Sheets credentials under the victim workspace, and if blockId is also known, wires those credentials into a victim Typebot's integration block - effectively hijacking the integration without the workspace owner's knowledge. No public proof-of-concept code has been identified, but the attack is straightforward given knowledge of the state encoding format, which is described in the advisory.