migration-guide
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- EXTERNAL_DOWNLOADS (SAFE): The skill mentions '@anthropic-ai/claude-code' as a target for migration. This is a package from a trusted organization (Anthropic) and the reference is provided as a manual instruction for the user rather than an automated download or execution.
- PROMPT_INJECTION (SAFE): The skill uses conditional logic and structured assessment questions to manage its workflow. These are benign instructional techniques used to guide the user and do not attempt to bypass safety filters or extract system prompts.
- DATA_EXPOSURE (SAFE): No credentials or sensitive system file paths are targeted. The skill correctly limits its scope to common tool configuration files like '.cursorrules' and 'CLAUDE.md'.
- COMMAND_EXECUTION (SAFE): While the skill generates migration plans that include shell commands (e.g., install and rollback commands), it lacks any mechanism to execute these commands autonomously, serving only as a text generator.
Audit Metadata