The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the axe CLI via a third-party Homebrew tap (cameroncooke/axe). This involves downloading and executing installation scripts from an external GitHub repository.
[COMMAND_EXECUTION]: The skill makes extensive use of shell commands to control the iOS simulator, including axe, xcrun, and ffmpeg. It also provides automation scripts that execute sequences of these commands.
[INDIRECT_PROMPT_INJECTION]: The skill ingests UI hierarchy data from the iOS Simulator, which represents an attack surface for indirect prompt injection if the app being tested contains malicious accessibility labels.
Ingestion points: The output of the axe describe-ui command is read by the agent and stored in files like accessibility-audit.txt as shown in SKILL.md.
Boundary markers: None are defined to separate UI metadata from agent instructions.
Capability inventory: The skill has the capability to execute shell commands, record video, and write to the local file system.
Sanitization: The instructions do not include steps to sanitize or validate the content of the UI elements retrieved from the simulator.

ios-app-tester