clickhouse-hello-world
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a standard educational guide for creating tables, inserting data, and querying a ClickHouse database using the official
@clickhouse/clientlibrary. - [CREDENTIALS_UNSAFE]: The skill demonstrates best practices by instructing the user to store sensitive connection details (host, username, and password) in environment variables instead of hardcoding them in scripts.
- [INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface where it processes data returned from database queries. However, it uses structured JSON parsing which significantly reduces the risk of the agent misinterpreting data as instructions.
- Ingestion points: Query results processed in
SKILL.md(Step 3) andreferences/system-tables.mdviaclient.query. - Boundary markers: Uses structured JSON parsing (
rs.json) to handle rows. - Capability inventory: Database schema creation, data insertion, and analytical querying.
- Sanitization: The skill uses typed JSON objects to handle database output.
Audit Metadata