The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the global installation of the lim CLI tool via npm install --global lim to interact with the Limrun cloud services.
[DATA_EXFILTRATION]: The skill provides commands to upload local application bundles and archives to the vendor's cloud storage via lim asset push and lim ios sync for simulation and manual preview purposes.
[COMMAND_EXECUTION]: The skill executes multiple shell commands using the lim CLI to manage simulator instances, launch applications, and manipulate files within the app container.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data originating from the mobile application under test, which could be used to influence the agent's behavior.
Ingestion points: Data enters the agent context through the accessibility element tree (lim ios element-tree), application log streams (lim ios launch-app), and directory listings (lim ios ls).
Boundary markers: The instructions do not specify any delimiters or safety warnings for handling content extracted from the application UI or logs.
Capability inventory: The agent has capabilities to write and read files from the app container (push-file, pull-file), interact with the UI (tap, type), and establish network tunnels (tunnel).
Sanitization: There is no evidence of sanitization or validation of the text content retrieved from the simulator's element tree before it is used for decision-making.

limrun-ios-simulator