python-on-vercel
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- External Content Retrieval: The skill is designed to fetch lesson content, curriculum maps, and metadata from
vercel.com. This ensures the teaching assistant has access to the most current course materials. - Command Line Interaction: To verify progress, the skill instructs the agent to run various Vercel CLI and FastAPI commands, such as
vercel whoami,vercel --version, andfastapi dev. These are standard diagnostic tools used to confirm the student's local environment is correctly configured. - Dynamic Instruction Processing: The skill includes logic to fetch and follow instructions embedded within remote lesson files (
<agent-instructions>blocks). This mechanism allows the course to provide specific guidance for different stages of the curriculum. - Data Ingestion Surface: The agent performs source code checks on student project files, including
api/index.py,app/page.tsx, and configuration files likepyproject.toml. This allows the agent to provide context-aware feedback based on the student's actual implementation. - User Confirmation Guardrails: The instructions explicitly caution the agent against performing sensitive actions, such as authentication, project linking, or production deployments, without the student's explicit request, maintaining user control over the development environment.
Audit Metadata