agent-accessibility-expert
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided data via the
$ARGUMENTSplaceholder inSKILL.md. This represents an attack surface for indirect prompt injection if the input originates from untrusted sources, such as external web content being audited for accessibility. - Ingestion points: The
$ARGUMENTSvariable at the end of theSKILL.mdfile receives the content to be processed. - Boundary markers: There are no explicit delimiters (like XML tags or block quotes) or instructions to ignore embedded commands within the input data.
- Capability inventory: The skill is scoped to a
general-purposeagent, which typically has access to tools for file reading, writing, and potentially shell execution. - Sanitization: The instructions do not include any steps to sanitize or validate the content provided in
$ARGUMENTSbefore the agent acts upon it.
Audit Metadata