The Agent Skills Directory

[REMOTE_CODE_EXECUTION] (MEDIUM): The better-env validate command dynamically imports configuration files from the project directory (e.g., src/lib/*/config.ts). This behavior constitutes dynamic execution of local code, which could be exploited if an agent is tasked with running validation on an untrusted repository.
[COMMAND_EXECUTION] (MEDIUM): The skill utilizes external CLI tools (specifically the Vercel CLI) to perform operations such as pulling, adding, and deleting remote environment variables. This grants the agent the ability to execute shell commands that interact with production infrastructure.
[DATA_EXFILTRATION] (LOW): The core functionality of the skill involves reading and writing sensitive credentials in .env files. While legitimate for a configuration tool, this creates a data exposure risk if the agent is manipulated into directing secrets to insecure locations.
[PROMPT_INJECTION] (LOW): The skill possesses an indirect prompt injection surface. (1) Ingestion points: .env files and project modules (src/lib/*/config.ts). (2) Boundary markers: Absent. (3) Capability inventory: File writing, CLI execution (Vercel), and dynamic module importing. (4) Sanitization: None described for ingested data.

better-env