planetscale-safe-orchestrator
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a robust safety contract that defaults to read-only operations and explicitly defines change gates to prevent unauthorized mutations. It strictly follows a verification discipline and stops execution unless explicit, named approval is provided by the user.
- [SAFE]: All external documentation and API references (planetscale.com/docs/llms.txt, planetscale.com/docs/openapi.yaml) are retrieved from official PlanetScale domains. As these are vendor-owned resources, they are considered safe for the skill's operational context.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection by ingesting untrusted data from the user's application repository and external documentation. 1. Ingestion points: External documentation (llms.txt) and application repository codebase. 2. Boundary markers: Absent for these external text sources. 3. Capability inventory: pscale CLI operations, API calls, and read-only SQL queries. 4. Sanitization: No specific sanitization or delimiters are noted for the ingested text, though the agent is instructed to prioritize official documentation over user-provided guesses.
Audit Metadata