altinity-expert-clickhouse-connection
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use
clickhouse-clientfor executing SQL queries if an MCP server is unavailable. This is a standard operation for a database diagnostic tool. - [INDIRECT_PROMPT_INJECTION]: The skill processes data returned from ClickHouse system tables (e.g.,
system.clusters,system.asynchronous_metrics). While this represents a surface for indirect prompt injection if the database content is attacker-controlled, the skill targets system-level diagnostics which are standard for its primary purpose. - Ingestion points: SQL query results from ClickHouse system tables (
SKILL.md). - Boundary markers: None explicitly defined in the prompt templates.
- Capability inventory: Execution of shell commands (
clickhouse-client) and SQL statements. - Sanitization: No explicit sanitization of database output is mentioned before processing.
Audit Metadata