The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads external schema files (SQL, TypeScript, Prisma) which serves as an ingestion point for untrusted data. Maliciously crafted schema definitions or comments could influence the generated seed data or the resulting script logic. Evidence:\n- Ingestion points: Reads all schema files identified via glob/grep patterns.\n- Boundary markers: No explicit delimiters or instructions to ignore embedded content are provided.\n- Capability inventory: The skill utilizes Write, Edit, and Bash tools.\n- Sanitization: No sanitization of ingested schema content is performed before processing.\n- [DYNAMIC_EXECUTION]: The skill generates TypeScript and SQL scripts for database seeding that are intended for execution. Evidence: The workflow includes writing seed logic to new files (e.g., 'scripts/seed.ts') based on project schema modeling.\n- [COMMAND_EXECUTION]: The skill instructs the agent to prompt users to execute the generated code using CLI tools. Evidence: Workflow steps explicitly document run commands such as 'npx tsx scripts/seed.ts' and 'npx wrangler d1 execute'.

db-seed