The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches mobile app design data, metadata, and screen images from the official Appllama MCP server at mcp.appllama.io.
[COMMAND_EXECUTION]: The instructions direct the agent to interact with local development tools, specifically the iOS Simulator or Android emulator, to run and verify generated code.
[DYNAMIC_EXECUTION]: The skill's primary purpose involves generating mobile application code (screens and components) and executing it within a simulation environment for iterative testing (the "simulator loop").
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata from external mobile apps (such as names, descriptions, and flow information) which presents a theoretical surface for indirect prompt injection.
Ingestion points: External app profiles and screen metadata fetched via search_apps, get_app, and list_app_screens (SKILL.md).
Boundary markers: No explicit prompt boundaries or "ignore instructions" warnings are defined when processing the retrieved metadata.
Capability inventory: The skill has capabilities to write files to the local research/ directory and execute shell commands for mobile simulators (build-from-scratch.md).
Sanitization: There is no mention of escaping or sanitizing external strings before they are incorporated into the agent's context.

appllama-usage