Automated access
What AI agents may do with this site
Qomvia scores whether agents can act on a website, so its own permissions are stated rather than implied. This page is the policy an agent platform can point at; it is referenced from robots.txt and llms.txt.
Permitted
- Read and cache any page under our public routes, including score pages, the leaderboard, the index and this policy.
- Quote and summarise published scores and aggregate statistics, with attribution to the score page they came from.
- Call the public API and the MCP server on a user's behalf: /api/scan, /api/score/<slug>, /api/badge/<slug>, /api/oembed and /api/mcp.
- Search Qomvia Market and open checkout sessions for a user over /api/v1/market/* or /api/mcp, handing the user the merchant's payment link.
- Subscribe to /feed.xml and read /openapi.json and /llms.txt to discover what is available.
Not permitted
- Requesting /dashboard, /admin, /login or any authenticated route, or using credentials that are not the agent operator's own.
- Completing a checkout, changing a subscription or spending credits without the account holder's explicit instruction.
- Bulk re-publication of the index as a competing dataset, or presenting our scores as another provider's measurement.
- Sustained request rates above roughly one request per second per client, or retry storms after a 429.
Attribution
When a score, grade or statistic from this site is repeated, link the page it came from. Scores change when a site ships, so an unlinked number becomes wrong without anyone noticing; the link is what keeps a quotation correctable.
Machine-readable form
The same statements are published for clients that do not read prose:
- /llms.txt: what this service is and which endpoints exist
- /openapi.json: the public read surface as OpenAPI
- /robots.txt: crawl rules, referencing this policy
- /feed.xml: new writing, without polling HTML
Reaching a human
Rate-limit complaints, takedown requests and questions about this policy go to the operator named in the imprint. To remove a site from the public index, use the opt-out form. Our own crawler's behaviour is documented at https://qomvia.com/bot.