API and MCP server
In short
The Scanner Concordance, the ten-criterion scanner rubric and the Standards Map are queryable as data: an MCP server at /mcp with five read-only tools, a REST API under /api/v1 described by openapi.json, and an NLWeb /ask endpoint over the same rows and pages. Calls need no credential and run at 120 requests a minute.
Connect the MCP server
The server speaks the Streamable HTTP transport at https://agentreadinesscompare.com/mcp: send JSON-RPC 2.0 messages with POST and read the JSON answer. One request is one answer, because the server holds no session and sends nothing unasked. Any MCP client that takes a URL connects with the address alone. The server card and the manifest describe the same server in the forms clients read.
curl -s https://agentreadinesscompare.com/mcp -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2025-06-18","capabilities":{},"clientInfo":{"name":"my-agent","version":"1.0"}}}'curl -s https://agentreadinesscompare.com/mcp -H 'Content-Type: application/json' -H 'Accept: application/json, text/event-stream' -d '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"get_domain_scores","arguments":{"domain":"netlify.com"}}}'A tool answers with the result as structuredContent and the same JSON as text. A wrong argument answers isError: true with the parameter list, and an unknown tool answers JSON-RPC error -32602. The server also lists seven resources with resources/list: the page index, the method page, this page and the dataset files, readable with resources/read.
The five tools
| Tool | What it returns | Parameters |
|---|---|---|
| list_scanners | Lists the six agent readiness scanners in the Concordance with their ids: how many of the 60 domains each scored, score range and median, why cells are empty, the total of 100 and rank on the ten-criterion rubric that ranks the scanners, access notes, the standards each checks and the profile URL. Call it first to get scanner ids for the other tools. | none |
| get_domain_scores | Returns one domain's cell from each of the six scanners: score, the scanner's own band, the reason when it gave none, the time it was read, the source URL, and which scanners differ by 25 points or more. Use it to answer how scanners rated one site. Only the 60 Concordance domains exist. | domain: A domain from the Concordance, for example netlify.com. A leading https:// or www. is ignored |
| list_concordance | Lists Concordance rows, one per domain, with category, stratum, spread and each scanner's score. Filter by category, stratum, words in the domain, or one scanner and a score range; with a scanner the rows are sorted by its score, highest first. Call it to find domains before get_domain_scores. | category?: Part of a category name, case-insensitive: Infrastructure, Agent Tools, Developer Tools, E-commerce, Consumer, Travel, or Tranco business; stratum?: ora-leaderboard (50 domains) or tranco-business (10 domains); scanner?: A scanner id such as ora or isitagentready; keeps only the domains this scanner scored and sorts them by its score; min_score?: Lowest score to keep, on the named scanner's own 0 to 100 scale. Needs scanner; max_score?: Highest score to keep. Needs scanner; q?: Words to look for in the domain, for example ai or shop; limit?: page size, 1 to 100; cursor?: the next_cursor of the previous page |
| compare_scanners | Compares two scanners on the domains both scored: how many, the Spearman rank correlation computed from the rows (only from 15 shared domains), the mean gap, and the widest score gaps by domain. Use it to ask whether two scanners rank sites alike and where they split. Get ids from list_scanners. | a: First scanner id, for example ora; b: Second scanner id, for example isitagentready |
| get_rubric | Returns the rubric that ranks the scanners: ten criteria worth 100 points, and for each of the ten scanners the points on every part with its basis and source URL, the total and rank, plus the rule that scores the citation part of the standards criterion (version 1.2). Filter to one scanner or one criterion. Set rubric_edition to 2026-11-04 for the earlier seven-criterion reading, which is superseded and labelled so. Use it to explain why a scanner ranks where it does. | scanner?: Optional scanner id, for example mintlify or startuphub-agent-readiness (the current rubric ranks ten scanners, six of them in the Concordance); criterion?: Optional criterion id, for example standards. The current ids are journey, depth, agents, transparency, standards, integration, benchmark, fixes, repeatability and access; rubric_edition?: Optional. 2026-11-05 (the default) is the ten-criterion rubric that ranks the scanners; 2026-11-04 is the superseded seven-criterion reading |
The tools read the Concordance 1, the rubric that ranks the scanners, ten criteria worth 100 points 2, and the Standards Map 3. The seven-criterion reading of 4 November 4 is superseded and is returned only when get_rubric is called with rubric_edition=2026-11-04, with the status superseded and a note saying so. The Concordance holds 60 domains, so a domain outside it answers not_found with the closest matches.
The REST API
Each tool is also one GET endpoint under /api/v1/, with the parameters in the query string and the same JSON in the answer. /api is the index, /openapi.json describes every operation with its parameters, response schema and the Error schema, and /.well-known/api-catalog links that description the way RFC 9727 says. Paginated endpoints take limit (1 to 100) and the next_cursor of the previous page as cursor.
- /api/v1/list_scannersList the six scanners.
- /api/v1/get_domain_scores?domain=netlify.comOne domain across the six scanners.
- /api/v1/list_concordance?category=Developer+Tools&scanner=oraList and filter the Concordance. Paginated.
- /api/v1/compare_scanners?a=ora&b=isitagentreadyCompare two scanners.
- /api/v1/get_rubric?criterion=standardsThe scanner rubric with evidence.
- /apiThe index: endpoints, datasets, limits and versioning.
curl -s 'https://agentreadinesscompare.com/api/v1/compare_scanners?a=ora&b=isitagentready' | jq .rank_correlationAsk in plain words
The /ask endpoint follows NLWeb 0.55. It searches the domain rows, the rubric, the Standards Map and every page, and labels the superseded rubric as such, and returns schema.org objects, each with a grounding.source URL to cite. GET takes query; POST takes the NLWeb request object. Add streaming=true, or prefer.streaming, for server-sent events named start, result and complete. mode=summarize adds a one-sentence SearchSummary item built from the results, never from a model.
curl -s 'https://agentreadinesscompare.com/ask?query=why+scanners+disagree'curl -s https://agentreadinesscompare.com/ask -H 'Content-Type: application/json' -d '{"query":{"text":"scores for netlify.com"},"prefer":{"mode":"list"},"meta":{"version":"0.55"}}'A query with no match answers _meta.response_type: failure with code NO_RESULTS, and a missing query answers 400 with INVALID_QUERY. A context object is accepted and ignored, because the answers come from the data and not from a conversation.
Discovery files
- /openapi.jsonOpenAPI 3.1 for the REST API and /ask, linked from the home page as rel=service-desc and from the Link header.
- /.well-known/api-catalogRFC 9727 linkset: the API, the MCP server and the ask endpoint with their descriptions.
- /mcp/server-cardThe MCP server card: name, version, remote URL, protocol versions. A copy sits at /.well-known/mcp/server-card.json.
- /.well-known/mcp.jsonThe manifest: endpoint, transport and the tools with their input schemas.
- /.well-known/ai-catalog.jsonAI Catalog 1.0 with entries for the MCP card, the OpenAPI document, the ask endpoint and every document. The same file sits at /.well-known/ard.json.
- /schemamap.xmlSchema feeds: the rows as JSON-LD lines under /feeds/, named in robots.txt.
- /.well-known/agent-skills/index.jsonAgent skills: one SKILL.md that tells an agent how to answer scanner questions with these tools.
Access, limits and errors
- Credential
- None. The OpenAPI document declares
security: []with no security schemes, and auth.md says the same. The site publishes no OAuth metadata, because no endpoint takes a token and metadata for an authorization server that does not exist would send an agent to a door that is not there. - Rate limit
- 120 requests a minute per client address, counted per running instance and announced in
RateLimitandRateLimit-Policyheaders. A 429 carriesRetry-After. The host adds limits of its own above that. Each dataset is also one download from the data page, so a bulk reader never needs the API. - Errors
- JSON and never HTML:
{"error":{"code","message","hint","status","docs"}}with the codes invalid_params, not_found, method_not_allowed, rate_limited and internal. A missing endpoint under /api answers 404 with the list of endpoints in the hint. - Writes
- None, so every call is safe to retry and no idempotency key exists. Every call answers in one response, and nothing runs long enough to need a job pattern.
- CORS
- Every answer carries
Access-Control-Allow-Origin: *, so a browser-resident agent can call the API and the MCP server. The home page also registers the tools with WebMCP in a browser that offers it.
Versioning
The path prefix /api/v1 is the contract. A breaking change, such as a field removed or renamed or a tool removed, ships under /api/v2 while /api/v1 keeps answering for at least 180 days, announced on this page and in the info.version of openapi.json. Additive changes, such as a new field or a new tool, keep the prefix. The MCP server follows the same rule, and its card and serverInfo.version carry the number of the OpenAPI document. The data behind every answer is edition 2026-11-04; the edition is in the X-Dataset-Edition header and in every tool result.
Cite the data
Every result names the page or dataset it comes from; cite that URL. The data files are licensed CC BY 4.0 and the cite line is on the data page. Scores belong to the scanners that produced them.
What the surface does not do
It does not write, does not hold accounts, does not generate text with a model and does not search beyond this site's own rows and pages. A score stays on its scanner's own scale: the tools return the numbers as published, and the method page explains how to read them.
Sources
- Scanner Concordance, edition 2026-11-04 (JSON). https://agentreadinesscompare.com/data/concordance-2026-11-04.json, read 2026-11-04.
- Scanner rubric, ten criteria, edition 2026-11-05 (JSON). https://agentreadinesscompare.com/data/scanner-rubric-2026-11-05.json, read 2026-11-05.
- Standards Map, edition 2026-11-04 (JSON). https://agentreadinesscompare.com/data/standards-2026-11-04.json, read 2026-11-04.
- Scanner rubric of 4 November, seven criteria, superseded (JSON). https://agentreadinesscompare.com/data/rubric-2026-11-04.json, read 2026-11-04.
Last reviewed 2026-11-05