db-migration-orchestrator

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to verify database connectivity using the dbmig Python package (python -m dbmig test-connection --side both). This is a core functional requirement for the migration pre-flight check.
  • [EXTERNAL_DOWNLOADS]: The skill prompts for the installation of Python dependencies via pip install -r scripts/requirements.txt. While the requirements file is local to the repository, the installation process will fetch packages from external registries.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from source database schemas and external configuration files (migration-config.yaml). 1. Ingestion points: Database metadata from source engines (Oracle/SQL Server) and user-provided configuration files. 2. Boundary markers: The skill enforces mandatory 'Human gates' at the end of every phase, requiring explicit user approval before moving to the next step. 3. Capability inventory: The skill performs file system writes to migrations/<project>/ and executes shell commands via the dbmig package. 4. Sanitization: The orchestrator relies on sub-skills and the underlying Python package for data processing; it does not define specific sanitization routines itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 02:18 AM