The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads the ByteHouse MCP server from the official Volcengine GitHub repository (github.com/volcengine/mcp-server) using the uvx tool at runtime. This is standard behavior for accessing the required diagnostic toolset.\n- [REMOTE_CODE_EXECUTION]: Executes the ByteHouse MCP server using uvx with a remote repository source. The source is a well-known repository maintained by the vendor, which aligns with the skill's intended functionality.\n- [COMMAND_EXECUTION]: Uses the uvx command to initialize and run the diagnostic server and scripts.\n- [DATA_EXFILTRATION]: The skill requests ByteHouse connection details (host, port, user, password) through environment variables. These credentials are used locally to authenticate with the database and are not transmitted to any unauthorized external services.\n- [PROMPT_INJECTION]: The skill ingests data from ByteHouse system tables (e.g., system.parts, system.mutations, system.query_log) for health reporting.\n
Ingestion points: Diagnostic data is retrieved from system tables via the run_select_query tool in scripts/cluster_diagnostics.py.\n
Boundary markers: No specific boundary markers or 'ignore' instructions are used for the ingested data.\n
Capability inventory: The skill can perform database queries, write JSON report files to the local output/ directory, and execute the MCP server command.\n
Sanitization: External database content is processed and formatted into reports without additional sanitization filters.

byted-bytehouse-diagnostics