The Agent Skills Directory

[DATA_EXFILTRATION]: The skill is configured to read sensitive configuration files including .env and .env.local. Although the instructions provide explicit negative constraints (anti-patterns) forbidding the printing or external transmission of these values, the inherent access to these sensitive paths constitutes a data exposure risk.
[COMMAND_EXECUTION]: The skill utilizes shell commands such as grep and git ls-files to inventory variables and verify the state of tracked files in the repository. These operations are performed locally and are consistent with the tool's stated purpose of environment auditing.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and parses untrusted source code from the repository to identify environment variable usage patterns.
Ingestion points: Scans various JavaScript, TypeScript, and configuration files throughout the codebase using recursive search tools.
Boundary markers: Absent; there are no specific instructions to the agent to treat content within the scanned files as data rather than instructions.
Capability inventory: The skill has the capability to read local files, execute shell commands, and modify project files like .env.example and .gitignore.
Sanitization: The skill uses a regular expression filter ([A-Z0-9_]{2,}) on the grep output to isolate variable names, providing a basic layer of filtering for the extracted data.

env-setup