The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Prisma CLI (npx prisma) and system utilities like ls, grep, and psql to detect environment settings and manage database operations. These are standard administrative actions for the primary purpose of the skill.\n- [INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection due to its analysis of user-provided project files.\n
Ingestion points: Reads and processes prisma/schema.prisma and files within the prisma/migrations/ directory.\n
Boundary markers: The instructions do not define markers to separate untrusted file content from agent instructions.\n
Capability inventory: The agent executes shell commands and database queries based on the project configuration.\n
Sanitization: No input validation or sanitization logic is specified for the ingested file contents.\n- [EXTERNAL_DOWNLOADS]: The skill uses npx prisma, which involves downloading packages from the npm registry, and links to official documentation at prisma.io. These resources are trusted and standard for this development context.\n- [CREDENTIALS_UNSAFE]: The documentation includes a placeholder DATABASE_URL with a sample password for configuration examples. This is an illustrative practice and not a security risk or exfiltration attempt.

prisma-expert