migration-architect
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external JSON data (migration specifications and database schemas) through its core scripts. While this provides a surface for indirect prompt injection, the scripts do not have access to sensitive credentials or external network resources, limiting the potential impact.
- Ingestion points: The scripts
migration_planner.py,compatibility_checker.py, androllback_generator.pyingest user-supplied JSON files via command-line arguments. - Boundary markers: The scripts rely on standard JSON parsing for structural validation but do not implement specific delimiters or warnings to prevent the agent from following natural language instructions embedded within the data.
- Capability inventory: The tools are restricted to local file read/write operations and structural data processing; no network exfiltration, arbitrary command execution, or sensitive file access was found in the code.
- Sanitization: Input content is parsed as structured data, and outputs are formatted as JSON or human-readable text reports.
Audit Metadata