Executive brief
dbt-mcp is a tool used to interact with dbt (data build tool) via the Model Context Protocol. A vulnerability was found where the software recorded sensitive information, such as raw SQL queries and credentials, into plaintext log files on the server's disk. If an unauthorized person gains access to these log files, they could view sensitive database queries and configuration details, potentially leading to further data exposure.
Technical details
A sensitive information disclosure vulnerability exists in dbt-mcp's logging mechanism. The `DbtMCP.call_tool()` function in `src/dbt_mcp/mcp/server.py` logs the raw `arguments` dictionary at the INFO level before tool execution and at the ERROR level during exceptions. When `DBT_MCP_SERVER_FILE_LOGGING` is set to true, these logs—containing unredacted `sql_query`, `vars`, and `node_selection` values—are written to `dbt-mcp.log` in plaintext. An attacker with local access to the log files could retrieve sensitive database interaction details or credentials passed via variables. The issue is fixed in version 1.17.1 by implementing argument redaction and improved input validation.
Affected products
- dbt-labs dbt-mcp < 1.17.1
Timeline
- 2026-05-05: patched: Fix committed and version 1.17.1 released.
- 2026-05-13: advisory: GitHub Security Advisory GHSA-7xgw-6qf3-7w59 published.
- 2026-07-16: disclosed: CVE-2026-44969 published to NVD.