awsflow-rdsdata

Pass

Audited by Gen Agent Trust Hub on Sep 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill enables dynamic SQL execution against Amazon RDS and Aurora Serverless clusters, which presents an attack surface for indirect prompt injection if external or untrusted data is processed.\n
  • Ingestion points: The sql and parameters fields in the ExecuteStatement and BatchExecuteStatement commands are populated based on user requests, which may include data from untrusted sources.\n
  • Boundary markers: The skill definition does not include explicit delimiters or instructions to the agent to disregard embedded instructions within the data provided to the SQL commands.\n
  • Capability inventory: The RDSDataTool has the capability to execute arbitrary SQL, perform batch updates, and manage database transactions (begin, commit, rollback).\n
  • Sanitization: While the tool structure supports parameterized queries, there are no specific instructions or logic within the skill to ensure the sql string itself is sanitized or to enforce the use of parameters over raw string concatenation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 5, 2026, 11:07 AM
Security Audit — agent-trust-hub — awsflow-rdsdata