arkweb-app-debug
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill includes shell, batch, and Python scripts (start-debug.sh, start-debug.bat, start-debug.py) that execute local commands such as hdc and a related utility script from the ohos-app-build-debug skill. These commands are used to detect the development environment, check for connected HarmonyOS devices, and establish port forwarding for debugging sessions.
- [INDIRECT_PROMPT_INJECTION]: The skill enables an AI agent to interact with and test ArkWeb applications (webviews) through the Chrome DevTools Protocol and MCP. This introduces a surface for indirect prompt injection if the web content being debugged contains malicious instructions designed to influence the agent's behavior.
- Ingestion points: External web content accessed through the DevTools connection, including the DOM, console logs, and network traffic (referenced in references/mcp-guide.md).
- Boundary markers: The skill does not implement explicit delimiters or instructions to the agent to ignore potentially malicious content within the webview data.
- Capability inventory: Through the referenced MCP server, the agent is granted capabilities to execute JavaScript (evaluate_script), automate inputs (click, fill), and monitor network activity.
- Sanitization: No sanitization or filtering of the webview content was identified in the provided scripts.
- [EXTERNAL_DOWNLOADS]: The documentation provides instructions for using chrome-devtools-mcp, which is an official and well-known tool from Google for bridging Chrome DevTools and the Model Context Protocol.
Audit Metadata