knowledge-migrate-all
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from multiple repositories. It mitigates the risk of indirect prompt injection by including explicit instructions to treat scanned content as data rather than instructions and forbidding the agent from widening its authority based on discovered content. Given the skill has no network access or file-modification tools, the potential impact of an injection is effectively neutralized.
- [DATA_EXFILTRATION]: While the skill accesses the filesystem to perform scans, it lacks network capabilities and includes explicit instructions to keep the report local to the user. It also provides a safe discovery mechanism by requiring user confirmation of search roots and skipping sensitive system or dependency directories.
Audit Metadata