Measure the AI shelf from your own tools.

Two read endpoints and an MCP server. Scores only — the measurement corpus and question graph stay ours. Free tier: 50 calls per key per day.

Get a key

Email hello@tryfrontrow.app with one line about what you're building. Keys look like frk_... and go in the Authorization header.

AI-readiness score

Can machines read this store? 0–100, with the checks. Add run=true to measure fresh.

curl "https://api.tryfrontrow.app/v1/readiness?domain=allbirds.com&run=true" \
  -H "Authorization: Bearer frk_..."

{"domain":"allbirds.com","score":88,"band":"ready", ...}

AI shelf share

The latest measured share of AI answers naming the brand, with confidence interval and the full report link. Available for domains on our measurement record.

curl "https://api.tryfrontrow.app/v1/score?domain=allbirds.com" \
  -H "Authorization: Bearer frk_..."

{"domain":"allbirds.com","shelf_share":0.75,"confidence_interval":0.02, ...}

MCP server

Give your agent the shelf. Streamable HTTP at https://api.tryfrontrow.app/mcp; tools: ai_readiness_score, ai_shelf_score.

// claude code
claude mcp add frontrow https://api.tryfrontrow.app/mcp \
  -t http -H "Authorization: Bearer frk_..."

Fair use: scores may be shown in your product with attribution ("Measured by FrontRow"). Reselling raw responses or bulk-harvesting the corpus is not licensed. Full audits and monitoring run on plans.