The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The seed.ts example in SKILL.md contains a hardcoded password 'admin123'. While used in a template for seeding development environments, hardcoded credentials pose a risk if used in production setups.
[INDIRECT_PROMPT_INJECTION]: The skill generates code for search functionality (getPublishedPosts in SKILL.md) that accepts a 'search' string. This creates a surface where malicious content retrieved from the database could influence the agent's logic if processed without proper isolation.
Ingestion points: The search parameter in the getPublishedPosts function in SKILL.md.
Boundary markers: None present in the generated query logic.
Capability inventory: The skill uses Read, Write, Edit, and Bash tools as defined in SKILL.md frontmatter.
Sanitization: Prisma's contains operator provides SQL-level protection but does not sanitize for indirect prompt instructions.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to run npx prisma commands for database migrations, client generation, and seeding tasks.
[DYNAMIC_EXECUTION]: The skill generates a seed script and executes it using tsx, and performs automated code generation via npx prisma generate to create the Prisma Client.

prisma-orm