CVE-2026-77329 in Tapo
Summary
by MITRE • 11/20/2026
Tapo provides Rust and Python APIs and an MCP server for TP-Link Tapo smart devices. From 0.3.1 until 0.5.0, the Streamable HTTP implementation in tapo-mcp/src/server.rs calls disable_allowed_hosts() without replacement Host or Origin validation, while TAPO_MCP_API_KEY remains optional for a listener that can bind to all interfaces. A malicious website can use DNS rebinding to send same-origin browser requests to the victim's local or LAN Tapo MCP listener when it runs without an API key. The attacker can invoke exposed tools including control_device for smart-device actuation and take_snapshot for configured cameras. Deployments with TAPO_MCP_API_KEY set are not affected by the unauthenticated path, and exploitation requires a running listener configured with real Tapo credentials and devices, plus victim interaction with the malicious site. This issue is fixed in version 0.5.0.
Once again VulDB remains the best source for vulnerability data.
Analysis
by VulDB Data Team • 11/20/2026
The vulnerability identified in TP-Link Tapo smart device management interfaces stems from an insecure configuration of the Streamable HTTP implementation within the tapo-mcp server component, specifically affecting versions ranging from 0.3.1 to 0.5.0. The core technical flaw lies in the invocation of disable_allowed_hosts() without implementing a replacement mechanism for Host or Origin header validation. This architectural decision effectively disables critical browser security controls that are designed to prevent cross-origin requests from unauthorized sources. When combined with the fact that the TAPO_MCP_API_KEY remains optional during listener initialization, the service becomes susceptible to unauthenticated access if no API key is explicitly configured by the administrator. The server's ability to bind to all network interfaces further exacerbates this risk, as it exposes the management interface not only to local loopback traffic but also to broader LAN segments and potentially external networks depending on routing configurations.
The operational impact of this vulnerability allows a malicious website to exploit DNS rebinding techniques to bypass same-origin policy restrictions in modern web browsers. By manipulating DNS responses during the execution of JavaScript code hosted on an attacker-controlled domain, the adversary can trick the victim's browser into sending HTTP requests that appear to originate from the trusted local or LAN IP address where the Tapo MCP listener is running. This technique effectively neutralizes standard cross-origin protections because the browser perceives the request as coming from a same-origin source after the DNS rebinding occurs. Consequently, an attacker can interact with exposed tools such as control_device for actuating smart devices and take_snapshot to capture images from configured cameras without providing any authentication credentials.
Exploitation of this vulnerability requires specific conditions to be met within the target environment. The victim must have a running instance of the Tapo MCP listener that is configured with real Tapo credentials and associated physical devices, as the API endpoints rely on these backend connections to perform actions. Furthermore, successful exploitation necessitates direct interaction from the victim, who must visit or interact with the malicious website containing the exploit code. Deployments where the TAPO_MCP_API_KEY environment variable is set are not affected by this unauthenticated path because the server enforces authentication checks regardless of host validation status. This distinction highlights that the vulnerability is primarily a configuration weakness rather than an inherent flaw in all deployments, though it remains critical for users who prioritize convenience over strict access control on their local networks.
From a classification perspective, this issue aligns with CWE-346 which describes the origin assumption error where software assumes a request originates from a trusted source without proper verification. It also relates to CWE-285 regarding improper authorization checks and CWE-918 for Server-Side Request Forgery (SSRF) variants that leverage DNS rebinding. In terms of the MITRE ATT&CK framework, this vulnerability facilitates techniques associated with Initial Access through browser-based attacks and potentially Discovery or Collection if camera snapshots are exfiltrated. The lack of proper origin validation is a common pitfall in web application development when developers disable default security features for debugging or convenience without implementing adequate compensating controls.
To mitigate this risk, administrators should ensure that the TAPO_MCP_API_KEY is always set to a strong, unique value before deploying the service. This enforces authentication on all requests, rendering DNS rebinding attacks ineffective as they cannot bypass credential-based access control even if they successfully spoof the origin header. Additionally, binding the listener exclusively to localhost or 127.0.0.1 prevents exposure to LAN-wide threats and external networks unless explicitly required for specific use cases that can be secured via network segmentation or firewall rules. Upgrading to version 0.5.0 or later is recommended as it addresses these implementation flaws by restoring proper host validation mechanisms. Organizations should also conduct regular audits of their IoT management interfaces to ensure that default configurations do not leave services exposed to unauthenticated access, particularly on networks where trusted and untrusted devices coexist.