The Agent Skills Directory

[SAFE]: The skill serves as a technical manual for the @prisma/composer ecosystem. It describes standard development workflows including project initialization, service configuration, and deployment to official Prisma infrastructure.
[DATA_EXPOSURE_&_EXFILTRATION]: The skill provides best practices for handling sensitive data. It introduces the SecretString type and the envSecret() function to ensure that credentials like PRISMA_SERVICE_TOKEN and SCHEDULER_API_KEY are managed via platform environment variables rather than hardcoded in source code. Secrets are redacted by the framework, which is a defensive security pattern.
[UNVERIFIABLE_DEPENDENCIES]: The skill references official Node.js packages from the @prisma scope (@prisma/composer and @prisma/composer-prisma-cloud). It mentions a plugin convention for extensions named prisma-composer-* on npm, which is a standard ecosystem pattern. Users are advised to verify package existence on npm before use.
[INDIRECT_PROMPT_INJECTION]: The skill describes an architecture that ingests user-defined schemas and code. It mitigates injection risks by enforcing schema validation through standard libraries (such as ArkType) and providing clear boundaries for service inputs.

prisma-composer