The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents functions that ingest data from potentially untrusted external surfaces.
Ingestion points: The skill references useUrlSearchParams.md, useClipboard.md, useStorage.md, and useFetch.md, which pull data from browser query strings, the system clipboard, browser storage, and network responses respectively.
Boundary markers: Usage examples in the documentation lack explicit boundary markers or instructions to ignore embedded commands.
Capability inventory: The skill documents capabilities such as useFetch and useAxios (network requests), useFileSystemAccess (local file interaction), and useScriptTag (dynamic JavaScript loading).
Sanitization: The provided usage examples do not demonstrate input sanitization or validation before processing external data.
[EXTERNAL_DOWNLOADS]: The documentation references and recommends the installation of various third-party libraries from official registries, including async-validator, axios, change-case, universal-cookie, drauu, focus-trap, fuse.js, idb-keyval, jwt-decode, nprogress, and qrcode.
[DYNAMIC_EXECUTION]: The skill documents the useScriptTag and useWebWorkerFn composables, which enable the application to load external scripts at runtime and execute functions within Web Workers, representing powerful execution capabilities.

vueuse-functions