Vercel's overall agentability score of 41/100 indicates moderate readiness for autonomous AI agent interaction. The homepage scored 41/100, pricing 31/100, and documentation 45/100, suggesting that agents will face friction when attempting to parse product information, evaluate costs, or navigate technical resources programmatically.

The audit identified three critical weak points: transparency scored 0/100, indicating no machine-readable confidence signals, audit logs, or source attribution for system outputs; shadow UI avoidance scored 30/100, pointing to interface elements that are difficult for agents to detect or interact with reliably; and defaults scored 31/100, meaning forms and inputs lack pre-filled sensible values that would accelerate agent-driven workflows.

Stronger performance in machine readability (66/100) and status communication (65/100) provides a foundation to build on, but the combination of poor transparency and weak defaults means agents currently struggle to operate Vercel's interfaces with confidence or efficiency.

Vercel can improve agentability by addressing the following issues identified in the audit:

• Add confidence indicators and source citations to generated outputs, such as a confidence field in API responses or 'verified vs best-guess' labels in the UI, and link outputs to their originating inputs or data sources.
• Implement a machine-readable activity log or 'why this result' feature that exposes system actions as both a visible feed and a JSON event log, enabling agents to audit and understand decision chains.
• Provide labeled pause/cancel controls for long-running operations with accessible names like 'Cancel deployment' rather than icon-only buttons, making agent control more reliable.
• Pre-fill form inputs with sensible defaults wherever reasonable values exist, reducing the decision load on agents and accelerating common workflows.
• Fix heading structure by using a single

  per page and maintaining contiguous hierarchy without skipped levels, improving document parsing accuracy.

• Rephrase section headings as questions they answer (e.g., 'How do I cancel my subscription?') to help agents quickly map content to user intent.