CVE-2026-10239: JeecgBoot Server-Side Request Forgery (SSRF) in Word Editing Module

Technical summary

The vulnerability exists in JeecgBoot's word editing functionality where the `WordUtil.addImage` method processes user-supplied image URLs without adequate validation or filtering. An authenticated user can craft a malicious request to the `/airag/word/edit` endpoint specifying an internal URL (such as localhost services, cloud metadata endpoints, or internal APIs) as the image source. The server will fetch this resource and process it, effectively becoming a proxy for the attacker. This is classified as CWE-918 (Server-Side Request Forgery). The CVSS v3.1 score of 6.3 reflects the requirement for authentication and the moderate impact on confidentiality and integrity.

Business impact

For organizations using JeecgBoot in production, this vulnerability creates multiple business risks. Attackers with valid credentials—or those who have compromised such credentials—can probe internal infrastructure, extract sensitive configuration data, or access internal APIs that should not be internet-facing. In cloud environments, SSRF vulnerabilities often lead to metadata endpoint compromise, potentially exposing cloud credentials or service tokens. Development teams may inadvertently expose database connection strings, API keys, or other secrets stored in internal systems. The public disclosure status increases the likelihood of exploitation attempts against unpatched instances.

Affected systems

JeecgBoot versions up to 3.9.2 are confirmed vulnerable. The vulnerability requires authentication to exploit, so it poses the greatest risk to environments where user accounts are broadly provisioned or where account compromise is feasible. Organizations running older versions of JeecgBoot in production environments, particularly those exposed to the internet or to untrusted networks, face the highest exposure. The module affected (`/airag/word/edit`) suggests this impacts deployments utilizing the word document processing features.

Exploitability

The vulnerability is remotely exploitable but requires a valid authenticated session. An attacker must either possess legitimate credentials or have successfully compromised a user account. The attack has no complex preconditions beyond authentication—the malicious request can be crafted through standard HTTP clients or integrated into automated attack workflows. Public disclosure of the vulnerability without an immediate patch increases the likelihood that exploit code or techniques are already circulating or being developed. The medium CVSS score reflects these authentication requirements; organizations with strong access controls and credential hygiene will face reduced risk.

Remediation

A fix is planned for an upcoming JeecgBoot release, though it has not yet been published at the time of this advisory. Immediately, organizations should implement access restrictions on the `/airag/word/edit` endpoint, limiting it to trusted users and internal networks only. Verify which versions of JeecgBoot are deployed in your environment and prioritize patching once the vendor releases a patched version. In the interim, consider disabling word document editing features if they are not critical to operations, or deploy web application firewall (WAF) rules to block suspicious image URL patterns (e.g., internal IP ranges, localhost, cloud metadata services).

Patch guidance

Monitor the JeecgBoot project repository and official release channels for version 3.9.3 or later. Once available, apply patches promptly, especially for internet-facing instances. Test the patch in a staging environment to ensure compatibility with existing workflows before production deployment. Verify that the patched version properly validates and restricts image URLs to legitimate external sources and rejects internal IP addresses, localhost references, and cloud provider metadata endpoints.

Detection guidance

Monitor HTTP requests to `/airag/word/edit` endpoints for suspicious patterns: requests containing localhost URLs, private IP ranges (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16), cloud metadata endpoints (e.g., 169.254.169.254 for AWS), or unusual domain names. Enable detailed logging of the `WordUtil.addImage` function to capture the actual image URLs being processed. Correlate unusual outbound HTTP requests from the JeecgBoot application server with inbound requests from users—a spike in internal probing activity may indicate exploitation. Network segmentation and egress filtering will limit the damage even if exploitation occurs.

Why prioritize this

Although the CVSS score is MEDIUM, this vulnerability merits prompt attention because it combines remote exploitability, public disclosure, and potential for lateral movement or credential theft in internal networks. The authentication requirement prevents immediate mass exploitation of completely unpatched internet-facing instances, but organizations with weak access controls or widespread credential compromise should treat this as higher priority. The planned availability of a patch in an upcoming release provides a clear remediation path.

Risk score, explained

The CVSS v3.1 score of 6.3 (MEDIUM) reflects: network exploitability (AV:N), low attack complexity (AC:L), authenticated attacker requirement (PR:L), no user interaction needed (UI:N), unchanged scope (S:U), and limited impact on confidentiality, integrity, and availability (C:L, I:L, A:L). The score is not elevated to HIGH because the vulnerability requires valid authentication and does not enable remote code execution or complete compromise. However, the score does not fully capture the risk in environments with poor credential security or the upstream impact potential in cloud or complex network topologies.

Frequently asked questions