exasol-notebook-connections

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute multiple Python scripts (check_backend.py, open_pyexasol.py, open_sqlalchemy.py, etc.) to validate connection states and perform database smoke tests.
  • [INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection by accepting and forwarding keyword arguments to underlying database and storage clients.
  • Ingestion points: User-provided parameters for connection helpers (**kwargs in references/connections-api.md) and configuration values retrieved from ai_config.db (SKILL.md).
  • Boundary markers: None present in instructions or script templates to distinguish between instructions and data.
  • Capability inventory: Execution of SQL queries via PyExasol, SQLAlchemy engine connection, and Ibis; file upload/download and UDF path resolution via BucketFS helpers (scripts/open_pyexasol.py, scripts/open_bucketfs.py).
  • Sanitization: The provided templates do not include explicit validation or sanitization logic for arguments passed to the underlying connection libraries.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 04:31 PM
Security Audit — agent-trust-hub — exasol-notebook-connections