Migrate

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly instructs the agent to read arbitrary pasted or file content and "display full content (not just preview)" and commit with provenance, which forces the LLM to output user data verbatim and can expose API keys/passwords or other secrets contained in those files.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). SKILL.md’s /migrate required workflow ingests outsider-authored free text when the user supplies pasted content via stdin or points the tool at arbitrary .md/.txt files/directories, which are then scanned by MigrateScan.ts and classified by the agent.