agent-vue-nuxt-expert
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an insecure integration pattern for processing external user input.
- Ingestion points: The
$ARGUMENTSplaceholder at the end ofSKILL.mdserves as the entry point where untrusted user data is interpolated into the agent's context. - Boundary markers: Boundary markers are absent. The user input is appended directly after the '## Your task' header without delimiters (such as XML tags or triple dashes) or instructions to the model to treat the content as data rather than directives.
- Capability inventory: The agent is configured as 'general-purpose', granting it access to file system operations, shell command execution (e.g., via
nuxi,npm, andvitest), and network requests (e.g.,useFetch,$fetch). - Sanitization: No evidence of sanitization, escaping, or validation is provided to handle the user-supplied content before it is processed by the agent.
- [PROMPT_INJECTION]: The use of the
$ARGUMENTSplaceholder without clear delimiters creates a vulnerability where a user can provide input that overrides the skill's preceding instructions. This design flaw facilitates prompt injection attacks where user input can command the agent to disregard safety guidelines or perform restricted actions outside the scope of Vue/Nuxt expertise.
Audit Metadata