catalog-publishing

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the agent to execute npm run build to compile the Next.js catalog website. This is a standard project build command and is used within the context of the defined deployment workflow.
  • [EXTERNAL_DOWNLOADS]: The skill instructions involve deploying code to Vercel and creating repositories on GitHub. These are well-known technology services and their use here is consistent with the skill's primary purpose of publishing a web catalog.
  • [PROMPT_INJECTION]: The skill defines a workflow that ingests content from external files such as prompt.md and meta.yaml within the design-packs/ directory. This creates a surface for indirect prompt injection. Evidence Chain: 1. Ingestion points: catalog.json and files within design-packs/ (SKILL.md). 2. Boundary markers: None specified in the instructions. 3. Capability inventory: Execution of build scripts and interaction with external version control/hosting providers. 4. Sanitization: Limited to status filtering (status: pass) and manual QA checklists.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 10:03 AM
Security Audit — agent-trust-hub — catalog-publishing