CVE-2026-77259 in mcp-atlassianinfo

Summary

by MITRE • 09/22/2026

MCP Atlassian is a Model Context Protocol (MCP) server for Atlassian products (Confluence and Jira). Prior to 0.22.0, confluence_upload_attachment opens a caller-selected server-local file without checking that the resolved path remains in the workspace. A caller can upload environment files, credentials, or other readable host data to a Confluence page and retrieve it through Atlassian. The advisory traces the vulnerable input and processing flow through confluence_upload_attachment, file_path, and open(file_path, "rb"), which identify the affected entry points, controls, and code paths. This issue is fixed in version 0.22.0.

If you want to get best quality of vulnerability data, you may have to visit VulDB.

Analysis

by VulDB Data Team • 09/22/2026

The vulnerability identified in MCP Atlassian versions prior to 0.22.0 represents a critical path traversal flaw within the confluence_upload_attachment function of the Model Context Protocol server implementation for Atlassian products such as Confluence and Jira. This security defect stems from an insufficient validation mechanism when handling file paths provided by external callers or automated agents interacting with the MCP server. Specifically, the application accepts a user-supplied file_path parameter which is directly passed to the open function in binary read mode without performing any checks to ensure that the resolved absolute path remains within the designated workspace directory boundaries. This lack of canonicalization and boundary enforcement allows an attacker to manipulate the input string using standard directory traversal sequences, such as dot-dot-slash patterns or encoded variations thereof, to escape the intended sandboxed environment.

From a technical perspective, this flaw aligns with CWE-22: Improper Limitation of a Pathname to a Restricted Directory, commonly known as path traversal. The core issue lies in the failure to resolve the file path against the current working directory or workspace root and verify that the resulting absolute path starts with the expected prefix. By bypassing this control, an adversary can reference arbitrary files on the host operating system where the MCP server is executing. This capability transforms a seemingly benign file upload operation into a powerful data exfiltration vector, as the attacker gains read access to any file for which the process running the MCP server has permissions.

The operational impact of this vulnerability is severe due to its potential for sensitive information disclosure and credential theft. An authenticated or otherwise authorized caller can exploit this flaw to upload critical system files from the host environment directly into a Confluence page, thereby making them accessible through the Atlassian interface. Typical targets include environment configuration files such as .env, which often contain database passwords, API keys, and secret tokens; SSH private keys used for server authentication; cloud provider credentials stored in local profiles; and other sensitive data residing on the filesystem. Once these files are uploaded to Confluence, they become part of the collaborative workspace content, potentially exposing them to a broader set of users who have access to that space or page, significantly expanding the blast radius beyond just the initial attacker.

This attack scenario maps directly to several techniques within the MITRE ATT&CK framework. The exploitation of file path manipulation corresponds to T1083: File and Directory Discovery, as it involves locating and accessing specific files on a system. Furthermore, since the data is exfiltrated via an application layer protocol (HTTP/HTTPS) through Confluence, it also aligns with T1567: Exfil Over Web Service, where stolen data is uploaded to external servers or services controlled by the attacker under the guise of legitimate file uploads. The ability to read arbitrary host files can serve as a stepping stone for further lateral movement if those credentials are leveraged against other systems within the enterprise network.

To mitigate this vulnerability, organizations must immediately upgrade MCP Atlassian to version 0.22.0 or later, where the developers have implemented proper path validation logic. This fix ensures that any file requested for upload is canonicalized and verified to reside strictly within the allowed workspace directory before being opened by the application. In environments where upgrading is not immediately feasible, defensive measures should include restricting the permissions of the user account running the MCP server to limit read access to sensitive directories such as home folders or system configuration paths. Additionally, implementing strict input validation on all file path parameters at the API gateway level can provide an additional layer of defense against path traversal attempts by rejecting requests containing suspicious characters or patterns indicative of directory escape sequences.

Responsible

GitHub M

Reservation

08/20/2026

Disclosure

09/22/2026

Moderation

accepted

CPE

ready

EPSS

0.00000

KEV

no

Activities

very low

Sources

Want to know what is going to be exploited?

We predict KEV entries!