altinity-expert-clickhouse-overview
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs diagnostic SQL queries against ClickHouse system tables like system.metrics and system.disks. This behavior is consistent with the skill's purpose for database health monitoring.
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data by reading from system logs and error tables.
- Ingestion points: The skill reads from system.text_log, system.query_log, and system.errors.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to treat this data as untrusted or to ignore instructions embedded within logs.
- Capability inventory: Query results are used to generate reports and route the agent to further specialist modules like altinity-expert-clickhouse-memory or altinity-expert-clickhouse-storage.
- Sanitization: The skill does not visibly sanitize log messages or error strings before they are presented to the agent, which could allow maliciously crafted log entries to influence the agent context.
Audit Metadata