vite
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill utilizes strong behavioral override markers including "CRITICAL", "MUST", and "Do not proceed without" to instruct the agent to ignore its internal training data in favor of external content.
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch information from
https://vite.dev/llms.txt. This targets the official domain of Vite, a well-known technology service. - [INDIRECT_PROMPT_INJECTION]: The skill establishes a vulnerability surface by mandating the ingestion of external documentation which the agent is instructed to use for verification and code generation.
- Ingestion points: Data enters the context via
WebFetchof thellms.txtfile inSKILL.md. - Boundary markers: The skill lacks explicit delimiters or instructions for the agent to ignore potential instructions embedded within the fetched text.
- Capability inventory: The agent is expected to perform file configuration, plugin setup, and environment variable management based on the external data.
- Sanitization: There are no instructions to sanitize or validate the content retrieved from the remote URL.
Audit Metadata