planetscale-schema-recommendations-agent-loop
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of purely instructional content with no embedded scripts, executables, or remote dependencies. It adheres to security best practices by implementing a policy of least privilege regarding database modifications.
- [PROMPT_INJECTION]: Analysis of indirect prompt injection surfaces found no significant risk. The skill processes external schema recommendations and application repository content, which constitutes an attack surface, but mitigates this via strict safety checks and mandatory human review.
- Ingestion points: External schema recommendation metadata (DDL, evidence from Insights) and application repository content.
- Boundary markers: Absent.
- Capability inventory: Codebase search, branch creation, and pull request generation.
- Sanitization: Mitigated by the 'Safety checks before proposing implementation' section and the 'Default output' policy requiring human approval for all migrations.
Audit Metadata