vueuse
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill provides implementation guides for utilities that ingest untrusted external data, creating an attack surface for indirect prompt injection.\n
- Ingestion points: The skill documents the use of
useFetch,useAxios,useClipboard,useIpcRenderer, anduseFileSystemAccessto handle external data (files:useFetch.md,useAxios.md,useClipboard.md,useIpcRenderer.md,useFileSystemAccess.md).\n - Boundary markers: Usage documentation lacks instructions for implementing delimiters or "ignore" instructions for external content when processing these inputs.\n
- Capability inventory: Documentation provides the agent with the ability to perform network operations, access the system clipboard, read/write local files, and communicate with system-level IPC.\n
- Sanitization: Instructions do not explicitly mandate sanitization of data before processing or interpolation into prompts.\n- [EXTERNAL_DOWNLOADS]: The documentation includes instructions for installing well-known third-party libraries from the npm registry to support specific integrations.\n
- Evidence: Installation commands for libraries such as
axios,focus-trap,qrcode, andidb-keyvalare found in the integration reference files.\n- [REMOTE_CODE_EXECUTION]: The skill documents functions capable of dynamic script execution and communication with high-privilege system environments.\n useScriptTag: Listed inSKILL.mdand the Browser section as a utility to dynamically load and execute JavaScript via script tags.\nuseIpcRenderer: Documents communication with the Electron main process, which operates outside the browser sandbox and can interact with system-level APIs (file:useIpcRenderer.md).
Audit Metadata