limrun-ios-simulator
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the
limCLI tool globally usingnpm install --global lim. This package is the required interface for the agent to communicate with the cloud simulator environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by processing external data.\n - Ingestion points: The agent reads the accessibility element tree, application logs, system syslog, and network traffic (HAR files) from the running iOS app and simulator.\n
- Boundary markers: The skill does not provide specific boundary markers or instructions to treat data from the simulator as untrusted when it is brought into the agent's context.\n
- Capability inventory: The agent has the ability to tap and type into the application, configure network tunnels, transfer files to the simulator, and launch or terminate processes.\n
- Sanitization: There are no sanitization or filtering steps mentioned for the data ingested from the simulator.\n- [DATA_EXFILTRATION]: The skill includes commands to upload local files to external storage (
lim asset push) and to capture network traffic from the simulator into local files (lim ios tunnel --har). While these are functional features of the service, they allow for data movement to external or local storage under the agent's control.\n- [COMMAND_EXECUTION]: The skill relies on executing various subcommands of thelimCLI to drive the simulator, manage instances, and handle networking tasks.
Audit Metadata