prisma-next-upgrade

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly requires refreshing/installing the unpinned remote skill package (Step 0: "pnpm dlx skills add prisma/prisma-next/skills/upgrade --all" and the "upgrade-skill subpath is intentionally unpinned (always main)"), and the agent is expected to read and execute the fetched per-transition upgrade instructions and colocated scripts (upgrades/*/instructions.md and scripts), so externally-hosted package content can directly influence tool actions and decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly tells the agent/user to (re)install/refresh from the unpinned repository reference "prisma/prisma-next/skills/upgrade" at runtime (pnpm dlx skills add prisma/prisma-next/skills/upgrade --all), and that fetched package contains the per-transition instructions and colocated scripts that directly control the agent's upgrade instructions and can be executed (e.g., ./stamp-storage-types-kind.ts, ./strip-inline-contracts.ts), so runtime fetching of that repo can change prompts/behavior and execute remote code.