The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to manage task lifecycles. This includes using jq to update state tracking files, sed for modifying plan documents, and standard file operations to create marker files.
[SAFE]: The skill follows security best practices for shell operations, such as using jq --arg to safely interpolate variables into JSON commands, which prevents command injection vulnerabilities.
[SAFE]: External interactions are restricted to invoking a subagent via the built-in Task tool and performing standard version control operations with git. No unauthorized network communication or data exfiltration attempts were found.
[SAFE]: File access is constrained to the specs/ directory and project-specific paths, following the principle of least privilege for task management.

skill-nix-implementation