NYC

find-hypertable-candidates

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (MEDIUM): The skill creates an attack surface by instructing the agent to ingest and analyze untrusted content from the database.
  • Ingestion points: Database query text from pg_stat_statements and schema metadata analyzed in SKILL.md.
  • Boundary markers: Absent; the agent is not provided with delimiters or instructions to treat query log content as untrusted data.
  • Capability inventory: The analysis output is used to rank tables for conversion, which influences subsequent decisions and potentially impacts database state via companion migration skills.
  • Sanitization: No sanitization or filtering of query log contents is performed before the agent processes them.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 11:06 AM