app-teardown

Pass

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the repository, including documentation and source code, to define its analysis parameters.
  • Ingestion points: Processes README, docs/PRD.md, DEFINE.md, DESIGN.md, package.json, and application source code files to understand the product and map screens.
  • Boundary markers: No explicit instructions are provided to the agent to distinguish between its own operational instructions and instructions that may be embedded within the repository's files.
  • Capability inventory: The agent is authorized to execute shell commands to install tools (Playwright), start local development servers, and interface with mobile simulators using xcrun simctl.
  • Sanitization: No sanitization or validation of the input documentation or code content is performed before it is used to guide the agent's teardown flow.
  • [COMMAND_EXECUTION]: The skill instructions require the agent to execute shell commands as part of the capture and environment setup process.
  • Evidence: The instructions explicitly direct the agent to start local development servers, install Playwright if missing, and capture screenshots using xcrun simctl io booted screenshot for iOS/Expo native applications.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 17, 2026, 04:07 AM