asset-prep
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted storefront data which could contain malicious instructions.
- Ingestion points: The skill reads external page source (lexsis-source.html) and compact manifests in SKILL.md.
- Boundary markers: The instructions do not specify boundary markers or include warnings to ignore instructions embedded within the ingested page content.
- Capability inventory: The skill performs file writes (page-manifest.json, page-preview.html) and uses tools for asset generation and importing (lexsis_drafts.asset_generate, lexsis_asset_upload.import).
- Sanitization: No sanitization or validation of the input storefront code is described.
- [EXTERNAL_DOWNLOADS]: The skill is configured to interface with a remote tool provider.
- Evidence: The skill uses an MCP server located at https://mcp.trylexsis.com/mcp to perform storefront media operations as defined in agents/openai.yaml.
Audit Metadata