The Agent Skills Directory

[DATA_EXFILTRATION]: The skill targets the creation and management of sensitive files including .env, .env.local, and .env.production.
Evidence: The skill description and instructions explicitly mention handling environment variables and configuration files which typically contain sensitive credentials like API keys and database connection strings.
Context: The risk is mitigated by the inclusion of .env.example templates and strict instructions to use .gitignore to prevent secret exposure in version control.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it involves reading and processing external data from configuration files.
Ingestion points: The agent is instructed to read content from .env files and configuration scripts.
Boundary markers: No specific instructions or delimiters are provided to the agent to treat the content of these files as untrusted data or to ignore embedded instructions.
Capability inventory: The skill allows the use of Bash, Write, and Edit tools, which could be exploited if an attacker-controlled configuration file is processed.
Sanitization: While the code examples include Zod for runtime validation, the agent's interaction with the files lacks explicit sanitization or instructions to ignore embedded commands.

environment-setup