accessibility-engineering
Pass
Audited by Gen Agent Trust Hub on Nov 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches and executes established accessibility auditing tools from the npm registry using npx.
- Evidence: Pinned versions of
@axe-core/cli@4.13.0andpa11y@9.1.1are referenced inSKILL.mdandscripts/audit-a11y.sh. - [COMMAND_EXECUTION]: The skill uses a bash script to coordinate the audit process, involving shell commands and local file operations.
- Evidence: The
scripts/audit-a11y.shscript manages temporary directories, executes CLI tools, and uses thenoderuntime to aggregate reports. - [DYNAMIC_EXECUTION]: The audit script executes JavaScript snippets via the Node.js CLI to process and merge JSON data.
- Evidence: Several
node -ecommands are used inscripts/audit-a11y.shto parse audit results and generate final reports. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external URLs, which could theoretically contain instructions intended to influence the agent.
- Ingestion points: Target URL content is fetched and analyzed by axe-core and pa11y in
scripts/audit-a11y.sh. - Boundary markers: The instructions do not define specific delimiters for separating violation data from the agent's core instructions.
- Capability inventory: The skill has access to shell execution and local file writing through its provided bash script.
- Sanitization: While the script verifies that tool outputs are valid JSON, it does not explicitly sanitize the text content of identified violations.
Audit Metadata