The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions to download source code from an external repository at https://github.com/tddworks/baguette and execute a make command, which compiles and runs code from a third-party source.
[EXTERNAL_DOWNLOADS]: The skill directs the user to install a software package using a third-party Homebrew tap (brew install tddworks/tap/baguette), which introduces dependencies from an unverified external source.
[COMMAND_EXECUTION]: The skill requires administrative privileges via sudo xcode-select to modify system-wide development environment settings.
[COMMAND_EXECUTION]: The skill launches a local web server on port 8421 (using baguette serve), which creates a network-accessible interface for controlling simulator devices.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection through its data processing and simulator management capabilities.
Ingestion points: The skill ingests untrusted data from simulator metadata via baguette list and processes external gesture sequences through baguette input.
Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands when processing JSON input from stdin or WebSocket streams.
Capability inventory: The skill manages system processes and simulator states via the baguette CLI and shell-based lifecycle commands.
Sanitization: The documentation provides no evidence of input validation or sanitization for the coordinate data and JSON commands received from external sources.

baguette-ios-simulator