audit-and-fix
Pass
Audited by Gen Agent Trust Hub on Apr 27, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill transmits HTML content, code snippets, and URLs to the AccessLint service via the
mcp__accesslint__*toolset. This is the intended behavior of the vendor's accessibility auditing functionality.- [COMMAND_EXECUTION]: The skill utilizes file system tools (Edit,Write) to apply code changes to the local repository based on recommendations received from the external audit service.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection. Because it processes data from external URLs and HTML files to determine source code edits, an attacker could embed malicious instructions in the audited content to influence the resulting code changes. The skill mitigates this risk by instructing the agent to confirm the scope of edits with the user and to use verifiable hooks (likedata-testid) when locating source code.
Audit Metadata