altinity-expert-clickhouse-mutations
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- SAFE (SAFE): The skill contains no executable code (Python, Node.js, or Shell scripts). It exclusively provides SQL queries meant for database diagnostics.
- Indirect Prompt Injection (LOW):
- Ingestion points: The skill reads data from ClickHouse system tables, specifically the
commandandlatest_fail_reasoncolumns insystem.mutations(seechecks.sql). - Boundary markers: None identified.
- Capability inventory: The skill only performs read-only diagnostic queries and provides recommendations; it has no network or shell execution capabilities.
- Sanitization: None, but since the output is analyzed by the agent for performance tuning rather than being re-executed, the risk of instruction injection is negligible.
- Data Exposure (LOW): The skill accesses database metadata such as table names, part counts, and mutation logs. This is appropriate and necessary for its primary purpose of database performance analysis.
Audit Metadata