byted-bytehouse-mcp
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the official ByteHouse MCP server from the Volcano Engine (ByteDance) GitHub repository using the
uvxtool. This is a verified resource provided by the authoring organization. - [COMMAND_EXECUTION]: The skill implements tools to execute arbitrary SELECT, DML, and DDL queries on connected ByteHouse clusters, which allows the agent to read, modify, or delete database content and schemas.
- [INDIRECT_PROMPT_INJECTION]: The skill processes database metadata and query results which serve as a surface for potential indirect prompt injection from untrusted database content. * Ingestion points: Database query outputs and table schema metadata. * Boundary markers: None identified in the client-side scripts. * Capability inventory: Arbitrary SQL execution, network communication with ByteHouse, and local file writes to the
output/directory. * Sanitization: The skill does not perform sanitization on database content or user-supplied SQL queries before they enter the agent context.
Audit Metadata