altinity-expert-clickhouse-index-analysis
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill extracts and analyzes historical queries from the
system.query_logtable, which can include attacker-controlled input if they have access to execute queries on the ClickHouse instance. - Ingestion points: Data is pulled from
system.query_login bothSKILL.mdandchecks.sql. - Boundary markers: There are no explicit delimiters or instructions to the agent to treat the data as untrusted or to ignore embedded instructions within the captured SQL.
- Capability inventory: The skill facilitates the execution of diagnostic SQL and analysis of the results to recommend schema changes.
- Sanitization: No sanitization is performed on the results of the
extractAllornormalizeQueryfunctions before presenting them to the LLM.
Audit Metadata