The Agent Skills Directory

Command Execution: The skill utilizes several command-line utilities, including dotnet, adb, xcrun simctl, and the maui devflow CLI. These tools are employed to manage builds, deployments, and device interactions, which aligns with the skill's primary function as a development aid.
Privilege Escalation: In references/linux.md, the skill provides instructions for configuring Linux UI automation tools that involve the use of sudo. Specifically, it describes installing system packages (xdotool, ydotool) and creating udev rules for the /dev/uinput device to grant persistent permissions to the user. While these are necessary steps for the tools to function, they involve the modification of system-wide configurations.
Indirect Prompt Injection Surface: The skill is designed to ingest and process data from the application under development and the host system, which could potentially include untrusted content.
Ingestion points: The skill scans project source files using grep (in SKILL.md and references/connectivity.md), retrieves Android system logs via adb logcat, and pulls files from devices (in references/android.md).
Boundary markers: There are no explicit markers or instructions to isolate or treat these data sources as untrusted when they are presented to the agent.
Capability inventory: The skill possesses capabilities including file system modification (rm -rf in references/ios-and-mac.md), network communication via the DevFlow agent, and execution of shell commands through the dotnet and maui CLIs.
Sanitization: There are no documented procedures for sanitizing or validating the contents of log files or visual tree data before they are analyzed by the agent.

maui-devflow-debug