The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell scripts (Bash and PowerShell) to identify the project root directory and detect current Git branch information. These scripts primarily read local configuration files like .adlc/init-options.json and project manifests (package.json, requirements.txt) to determine the appropriate tech stack.
[DATA_EXPOSURE]: The scripts read environment variables such as TEAM_AI_DIRECTIVES and local JSON configuration files to resolve pathing for the evaluation system. This data remains within the local project context and is not transmitted externally.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote downloads or execute code from external URLs. It initializes local Python grader scripts from internal templates to provide a security baseline for the user's project.

evals-init