Vercel Launches ‘Is Agentic’ Free Website Scoring Tool
Vercel has released Is Agentic, a free tool that evaluates public website readiness for AI agents using over 100 Ora checks and provides structured reporting.
Stock photo for illustration only, not from the actual event
- Vercel launched Is Agentic, a free agent-readiness scoring tool for public websites
- Utilizes over 100 checks from Ora with a grading scale ranging from A+ to F
- Available for free via web browser, read-only API, CLI, and MCP server without API keys
- Provides actionable recommendations and integrates directly into CI workflows
Vercel has introduced Is Agentic, a new tool designed to audit public websites and score their readiness for AI agents. The service is entirely free of charge with no paid plans, subscription fees, or per-report costs. Users can access the public site, read-only API, CLI, and MCP server immediately without requiring a billing account or an API key.
The tool caters to organizations of all types and sizes. Seed-stage startups can obtain a free baseline audit without procurement hurdles, mid-market SaaS teams can integrate JSON outputs into continuous integration pipelines as regression gates, and enterprise teams can benchmark developer portals and commerce surfaces across different business units.
Stock photo for illustration only, not from the actual event
Target industries span developer tools and SaaS, e-commerce and retail, travel and hospitality, fintech, marketplaces, healthcare directories, and media publishers where AI agents might shop, book, or cite information on behalf of users. Applications include pre-launch audits, CI checks that fail builds on server-rendered regressions, competitive benchmarking, and verifying the discoverability of MCP servers or OpenAPI surfaces.
Ora publishes a letter scale consisting of A+ for scores between 95 and 100, A for 86 to 94, B for 70 to 85, C for 48 to 69, D for 28 to 47, and F for 0 to 27. Vercel reorganizes these checks into its own displayed scoring model.
The scoring framework allocates an 80-point pool to essential checks and a 20-point pool to recommended checks, while emerging signals add a bonus capped at five points where absence never lowers a score. Non-applicable checks are excluded rather than counted as failures, and partial results receive proportional credit.
This applicability logic ensures that marketing sites are never unfairly penalized for omitting technical interfaces they never claimed to offer, resulting in a much fairer and more accurate assessment for diverse web platforms.
Every finding includes observed evidence and concrete recommendations. API responses supply structured data detailing the score, score label, scan timestamp, eligible checks, a score breakdown by tier, and an issue array containing identifiers, names, details, recommendations, results, and tiers. Real check identifiers include content-no-js, agent-friendly-404, markdown-negotiation-vary, json-ld, sitemap, trust-anchors, and metadata-completeness.
Completed reports are exposed through three read-only surfaces, including an API rate-limited to 120 requests per client IP per minute, and an MCP server running over Streamable HTTP. The platform also notes the background of data science professional Michal Sutter, who holds a Master of Science in Data Science from the University of Padova.
Source: MarkTechPost
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment