FastGPT — CraftedSignal Threat Feed

FastGPT NoSQL Injection Vulnerability (CVE-2026-40351)

hello@craftedsignal.io — Sat, 18 Apr 2026 12:00:00 +0000

FastGPT is an AI Agent building platform. Versions prior to 4.14.9.5 are susceptible to a critical NoSQL injection vulnerability (CVE-2026-40351) affecting the password-based login endpoint. The vulnerability stems from the use of TypeScript type assertion without runtime validation, enabling unauthenticated attackers to inject MongoDB query operators within the password field. This bypasses the intended password check, granting the attacker the ability to authenticate as any user, including the root administrator. Successful exploitation leads to complete control over the FastGPT instance and its associated data. This vulnerability was addressed in FastGPT version 4.14.9.5. All users of FastGPT versions prior to 4.14.9.5 are vulnerable to this attack.

Attack Chain

An unauthenticated attacker identifies a vulnerable FastGPT instance running a version prior to 4.14.9.5.
The attacker crafts a malicious HTTP POST request to the password-based login endpoint.
Within the POST request body, the attacker places a MongoDB query operator object (e.g., {"$ne": ""}) in the password field, bypassing the standard password check.
The vulnerable FastGPT application processes the malicious request without proper validation.
The MongoDB query operator is executed, bypassing the authentication mechanism.
The attacker is granted unauthorized access to the FastGPT application, assuming the identity of an arbitrary user, including the root administrator.
The attacker leverages their administrative privileges to access sensitive data, modify configurations, or perform other malicious actions within the FastGPT instance.

Impact

Successful exploitation of CVE-2026-40351 allows an unauthenticated attacker to gain complete control over a FastGPT instance. This can lead to unauthorized access to sensitive AI agent configurations, user data, and other critical information. The impact includes data breaches, service disruption, and potential compromise of downstream systems that rely on the FastGPT platform. Given the critical nature of AI agent building platforms, the compromise of a FastGPT instance can have far-reaching consequences.

Recommendation

Immediately upgrade all FastGPT instances to version 4.14.9.5 or later to patch CVE-2026-40351.
Deploy the Sigma rule Detect FastGPT NoSQL Injection Attempt to identify potential exploitation attempts targeting the login endpoint.
Monitor web server logs for unusual POST requests to the login endpoint, specifically looking for MongoDB query operators within the password field as detected by rule Detect FastGPT NoSQL Injection Attempt.
Review and restrict network access to the FastGPT instance to only authorized users and systems to minimize the attack surface.

FastGPT NoSQL Injection Vulnerability in Password Change Endpoint

hello@craftedsignal.io — Fri, 17 Apr 2026 22:16:32 +0000

FastGPT, an AI Agent building platform, is susceptible to a critical NoSQL injection vulnerability affecting versions before 4.14.9.5. The flaw resides within the password change endpoint, enabling an authenticated attacker to circumvent the necessary “old password” verification process. By injecting MongoDB query operators, an attacker with an existing, low-privileged session can manipulate password changes for their own account, or potentially other accounts if combined with ID manipulation techniques. This exploit leads to full account takeover, allowing attackers to maintain persistence and potentially compromise sensitive data. This vulnerability has been patched in version 4.14.9.5, urging users to upgrade immediately.

Attack Chain

Attacker gains initial access to a FastGPT account with low privileges through legitimate means (e.g., registration or stolen credentials).
Attacker navigates to the password change endpoint within the FastGPT application.
The attacker crafts a malicious request to the password change endpoint, injecting MongoDB query operators into the “old password” field. For example, using a payload like {$ne: "legitimate_old_password"}.
The application’s backend improperly processes the injected query operators, failing to correctly validate the old password against the stored hash.
The attacker provides a new password and confirms it within the crafted request.
The FastGPT application updates the account’s password in the database, replacing the original password with the attacker-controlled value.
The attacker logs out and logs back in using the newly set password, gaining full control of the compromised account.
The attacker leverages the compromised account to access sensitive data, modify configurations, or perform other malicious activities within the FastGPT platform.

Impact

Successful exploitation of this vulnerability allows attackers to take complete control of FastGPT accounts. The consequences range from unauthorized access to sensitive data and configurations to potential manipulation of AI agent behavior. This account takeover can lead to data breaches, service disruption, and reputational damage. While the specific number of victims is unknown, any FastGPT instance running a version prior to 4.14.9.5 is vulnerable, potentially affecting a wide range of users and organizations. The CVSS v3.1 base score of 8.8 highlights the severity of this issue.

Recommendation

Immediately upgrade all FastGPT installations to version 4.14.9.5 or later to patch the NoSQL injection vulnerability (CVE-2026-40352).
Implement the Sigma rule Detect FastGPT Password Reset Bypass to detect potential exploitation attempts against the password change endpoint.
Review FastGPT webserver logs for unusual patterns or MongoDB query operators within requests to the password change endpoint to identify potential compromises.
Enable and review detailed webserver logging for FastGPT to increase visibility into HTTP requests.

Critical Vulnerability in FastGPT Allows API Key Exfiltration and Internal Network Access

hello@craftedsignal.io — Wed, 01 Apr 2026 16:12:02 +0000

A critical vulnerability, CVE-2026-34162, has been identified in FastGPT, a framework for building AI-powered applications. The vulnerability resides in the HTTP tools testing endpoint, which is accessible without authentication. This allows an unauthenticated attacker to send arbitrary server-side HTTP requests and receive the responses. If the default admin token is not changed, an attacker can access the proxy management API to exfiltrate third-party API keys. Furthermore, the attacker can interact with and potentially exploit all Docker Compose internal services by manipulating HTTP headers. This issue was publicly disclosed on April 1, 2026, by CCB Belgium, who strongly recommends immediate patching. The vulnerability is patched in version 4.14.9.5. Successful exploitation can lead to complete control over the internal network and sensitive data exposure.

Attack Chain

An unauthenticated attacker identifies a vulnerable FastGPT instance exposed to the network.
The attacker accesses the FastGPT HTTP tools testing endpoint without authentication.
The attacker uses the endpoint to send arbitrary HTTP requests to the FastGPT server itself or internal services.
If the default admin token is unchanged, the attacker uses the HTTP proxy functionality to access the proxy management API.
The attacker exfiltrates third-party API keys stored within the FastGPT configuration.
The attacker leverages the exfiltrated API keys to access external services, potentially causing further damage.
The attacker uses the HTTP proxy functionality, including custom headers, to interact with other Docker Compose internal services.
The attacker exploits vulnerabilities in these internal services, leading to complete access to the internal network and sensitive data.

Impact

Successful exploitation of CVE-2026-34162 can lead to the complete compromise of the FastGPT server and the internal network it manages. An attacker can exfiltrate sensitive API keys, gain unauthorized access to internal services, and potentially pivot to other systems within the network. The vulnerability poses a high risk to the confidentiality and integrity of data, potentially impacting numerous organizations relying on FastGPT for their AI-powered applications. The CCB Belgium advisory highlights the potential for widespread impact given the nature of the vulnerability and the popularity of FastGPT.

Recommendation

Immediately patch FastGPT instances to version 4.14.9.5 to remediate CVE-2026-34162 as per the vendor advisory.
Implement the remediations documented in the vendor advisory to strengthen the security of FastGPT instances.
Upscale monitoring and detection capabilities to identify any related suspicious activity, ensuring a swift response in case of an intrusion, as recommended by the CCB.
Investigate and report any suspected intrusions using the incident reporting URL found in the advisory (https://ccb.belgium.be/report-incident).