The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/team-init.sh uses eval "$(resolve_team_levelup_paths)" to initialize environment variables. This pattern executes the output of a function as shell commands at runtime, which is a medium risk for local command execution if the helper function's output is manipulated.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze brownfield codebases to reverse-engineer patterns and generate Context Directive Records (CDRs).
Ingestion points: The skill reads arbitrary source code from project sub-systems (e.g., src/auth/, services/payment/) during the Discovery and Pattern analysis phases.
Boundary markers: No explicit boundary markers or instructions to ignore embedded malicious content within the scanned code are present in the skill definition.
Capability inventory: The skill has the ability to write to Git branches, execute shell scripts, and interact with the team-ai-directives repository.
Sanitization: There is no documented sanitization or filtering of the ingested codebase content before it is passed to the Discovery, Pattern, and Synthesis agents.
[COMMAND_EXECUTION]: The skill relies on shell scripts (scripts/team-init.sh) and PowerShell scripts (scripts/team-init.ps1) to perform setup, resolve paths, and interact with Git repositories. It specifically sources helpers from a relative parent directory (../team-levelup/scripts/helpers.sh), creating a dependency on external skill components.

team-init