The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a Python script (scripts/task_control.py) to handle logic for routing and task status. This script is part of the skill's own package and is executed to perform deterministic decisions based on local environment data. The command execution is constrained to this internal utility.
[DATA_EXPOSURE]: The skill implements a _safe_display function in scripts/task_control.py that uses regex patterns to detect and redact sensitive information such as API keys (OpenAI, GitHub), passwords, tokens, email addresses, and phone numbers before they are projected onto the taskboard UI.
[PATH_VALIDATION]: The skill employs rigorous path validation using os.realpath and os.path.commonpath to ensure that all tasks and workers are contained within host-verified 'allowed_roots'. This prevents path traversal and ensures the agent cannot interact with files outside the intended project scope.
[AUTHORIZATION_ENFORCEMENT]: The skill explicitly defines and limits authority profiles (implementation vs controlled-delivery). It prohibits high-risk operations like force pushing, history rewriting, or deployment by default, requiring separate explicit authorization for such actions.

workspace-taskboard