The Agent Skills Directory

[PROMPT_INJECTION]: The skill handles an indirect injection surface by processing external configuration files. Ingestion points: Local .env files loaded via the dotenv package. Boundary markers: None present. Capability inventory: Includes Read, Write, Edit, and Bash tools. Sanitization: The skill implements robust sanitization by recommending and providing code for runtime schema validation using the Zod library to ensure that all environment variables conform to expected types and formats.
[DATA_EXFILTRATION]: The skill manages sensitive configuration files such as .env and .env.production. It correctly identifies the risk of credential exposure and implements industry-standard mitigations by instructing the user to exclude these files from version control via .gitignore and providing a non-sensitive .env.example template.

environment-setup