ios-marketing-capture

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates and executes shell scripts that interact with system-level developer tools including xcodebuild, flutter, and xcrun simctl to build the app and capture screen output.
  • [EXTERNAL_DOWNLOADS]: The skill documentation provides instructions for installing the AppLaunchFlow MCP via npx from the NPM registry to enable optional layout generation features.
  • [REMOTE_CODE_EXECUTION]: The tool facilitates the execution of a remote MCP server using npx -y @applaunchflow/mcp@latest as part of the asset preparation pipeline.
  • [DATA_EXFILTRATION]: The skill is designed to upload curated application screenshots to the vendor's service (applaunchflow.com) via a specialized MCP tool to create professional App Store layouts.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its automated discovery mechanism.
  • Ingestion points: The agent is directed to read and analyze local project files including pubspec.yaml, project.pbxproj, and localization string files (.xcstrings, .arb).
  • Boundary markers: No explicit delimiters or instructions to treat project content as untrusted data are present in the scripts.
  • Capability inventory: The agent possesses the ability to execute shell commands, write files, and perform network requests via the integrated MCP.
  • Sanitization: No sanitization or filtering is applied to the data ingested from the user's codebase.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 02:09 PM