Executive brief
The mcp-sequential-thinking tool is a Model Context Protocol server used for AI-assisted task management with persistent session storage. A path traversal vulnerability in the import and export session functions allows an attacker to read, write, or delete arbitrary files on the system by manipulating the file_path parameter, potentially exposing sensitive data or compromising system integrity.
Technical details
This is a path traversal vulnerability in the import_session and export_session functions within mcp_sequential_thinking/server.py. The vulnerable code fails to properly validate or sanitize the file_path argument before using it in file operations, allowing an attacker to use directory traversal sequences (e.g., "../../../") to access files outside the intended directory. The attack is remotely exploitable over the network via the MCP protocol without requiring authentication. An attacker can read arbitrary files from the system, write malicious files, or delete existing files depending on the application's file permissions. The fix is available in version 0.6.0 and commit 2fad3ee8ab1d0868b6c1afb5895bc336a10e5267.
Affected products
- arben-adm mcp-sequential-thinking up to 0.5.0
Timeline
- 2026-08-28: disclosed: Vulnerability disclosed publicly
- 2026-06-13: patched: Fix merged in commit 2fad3ee, released in version 0.6.0