The Agent Skills Directory

[DATA_EXPOSURE_AND_EXFILTRATION]: The skill is designed to read and validate sensitive environment configuration files, including .env, .env.local, and other variants. Although the skill instructions explicitly forbid printing or transmitting secret values, the capability to read these files is an inherent part of its validation functionality and constitutes a data exposure surface.
[INDIRECT_PROMPT_INJECTION]: The skill performs a recursive search across the project's source code files (e.g., *.ts, *.js) to identify environment variable usage. This ingestion of untrusted codebase content as data creates a surface for indirect prompt injection, where malicious instructions hidden in code comments or strings could potentially influence the agent's behavior during the discovery phase.
[COMMAND_EXECUTION]: The skill utilizes shell commands such as grep and git ls-files to inventory environment variables and audit for tracked secrets. These commands are standard for development auditing tasks and are limited to the local repository scope.

env-setup