npm - agent-ready-mcp - Versions diffs - 0.7.0 → 0.7.2 - Mend

agent-ready-mcp 0.7.0 → 0.7.2

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Files changed (2) hide show

package/dist/mcp-server.mjs +24 -8
package/package.json +1 -1

package/dist/mcp-server.mjs CHANGED Viewed

@@ -31298,7 +31298,11 @@ The llmstxt.org spec treats some properties as foundational and others as option
 ## How are JavaScript-rendered pages handled?
-P23 detects pages where the static HTML response lacks data that only appears after client-side rendering \u2014 missing H1, empty body text, JSON-LD that injects after hydration. When P23 fires, the runner marks dependent checks (P10, P11, P12, P14) as \`unreliable\` and the scorer excludes them from both numerator and denominator. Without this, a single SPA architecture choice would compound into a 4-point score drop across unrelated checks.
+P23 detects pages where the static HTML response lacks data that only appears after client-side rendering \u2014 missing H1, empty body text, JSON-LD that injects after hydration. When P23 fires, the runner marks the dependent per-page checks (P12, P13, P14, P21) as \`unreliable\` and the scorer excludes them from both numerator and denominator. Without this, a single SPA architecture choice would compound into a multi-check score drop.
+**Only what exists is scored.** Most A-checks are conditional \u2014 "if the page has images, they need alt text" \u2014 so on a page with no images, no tables and no media there is nothing to grade. Those results are marked \`notApplicable\` and excluded from both numerator and denominator. Scoring them as passes rewarded a site for what it lacked. Checks that state a prohibition (no positive tabindex, no timed meta refresh) or a document-level requirement (a title, a language, a main landmark) always count \u2014 for those, absence is a real verdict.
+The same gate applies to the accessibility sub-score. Most A-series checks count elements and treat "found none" as a pass; on a client-rendered page the static DOM is a skeleton, so "found none" means "could not look". When P23 fails those checks are marked \`unreliable\` and dropped from the score; if every one is, the score is \`null\` rather than 0. The checks that read \`<head>\` or \`<html>\` (A9 viewport, A10 title, A13 lang, A17 meta refresh) stay trustworthy, because a JS-only shell still serves them.
 ## Where is the source for each check?
@@ -31407,7 +31411,7 @@ Discover-then-validate: when the relevant well-known endpoint returns 404, the c
 | C20 | AP2 payment protocol support |
 | C21 | ACP profile (/.well-known/acp.json) |
-## Accessibility checks (9)
+## Accessibility checks (23)
 Run over the homepage DOM (v1). WCAG-grounded accessibility-tree signals \u2014 image text alternatives, form labels, control names \u2014 plus a static layout-stability (CLS) proxy. Scored into a separate \`accessibilityScore\`, a distinct suite from the 69 checks above: accessibility is WCAG, not the Vercel Agent Readability Spec, so it never moves the Vercel score.
@@ -31422,10 +31426,23 @@ Run over the homepage DOM (v1). WCAG-grounded accessibility-tree signals \u2014
 | A7 | Heading hierarchy is well-formed |
 | A8 | No positive tabindex |
 | A9 | Pinch-zoom is not disabled |
-`;
+| A10 | Page has a title |
+| A11 | Accessible name contains the visible label |
+| A12 | Content is reachable past the chrome |
+| A13 | Language tag is valid |
+| A14 | Hidden elements are not focusable |
+| A15 | Interactive controls are not nested |
+| A16 | Headings are not empty |
+| A17 | No timed meta refresh |
+| A18 | Autocomplete tokens are valid |
+| A19 | Tables declare their headers |
+| A20 | SVG images have a text alternative |
+| A21 | Media declares a captions track |
+| A22 | List structure is well-formed |
+| A23 | ARIA roles and attributes are valid |`;
 var LLMS_TXT = `# Agent Ready
-> Agent Ready is a free tool that scores any website against the Vercel Agent Readability Spec, the llmstxt.org specification, and agent-protocol specs (MCP, A2A, agents.json). It runs 69 checks \u2014 plus a separate accessibility sub-score from 9 WCAG 2.2 / layout-stability checks \u2014 and provides actionable fix guidance for every failing check.
+> Agent Ready is a free tool that scores any website against the Vercel Agent Readability Spec, the llmstxt.org specification, and agent-protocol specs (MCP, A2A, agents.json). It runs 69 checks \u2014 plus a separate accessibility sub-score from 23 WCAG 2.2 / layout-stability checks \u2014 and provides actionable fix guidance for every failing check.
 This resource mirrors agent-ready.dev's own /llms.txt so MCP clients can introspect the same surface that ChatGPT, Perplexity, and other AI agents see when discovering Agent Ready as a tool.
@@ -31463,7 +31480,7 @@ Agent Ready's checks map to the specifications below. Each entry links to the ca
 ## Agent protocols
-- **MCP Server Cards (SEP-1649)** \u2014 The Model Context Protocol discovery card at /.well-known/mcp.json \u2014 its presence (C1) and required fields (C2). Canonical: <https://github.com/modelcontextprotocol/modelcontextprotocol/issues/1649> Checks: C1, C2.
+- **MCP Server Cards (SEP-2127)** \u2014 The Model Context Protocol discovery card at /.well-known/mcp.json \u2014 its presence (C1) and required fields (C2). SEP-2127 (a server.json subset) superseded the earlier SEP-1649 proposal. Canonical: <https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2127> Checks: C1, C2.
 - **OAuth Protected Resource Metadata (RFC 9728)** \u2014 Protected-resource metadata an MCP server advertises so agents can locate its authorization server. Canonical: <https://datatracker.ietf.org/doc/html/rfc9728> Checks: C3.
 - **A2A Protocol \u2014 Agent Cards** \u2014 The agent card at /.well-known/agent-card.json that lets agents discover and call other agents \u2014 existence with correct Content-Type (C4) and required fields (C5). Canonical: <https://a2a-protocol.org/v1.0.0/specification> Checks: C4, C5.
 - **Wildcard agents.json** _(pre-standard)_ \u2014 An OpenAPI extension declaring which existing REST endpoints agents should call. Pre-standard (v0.1.0). Canonical: <https://github.com/wild-card-ai/agents-json> Checks: C6.
@@ -31489,8 +31506,7 @@ Agent Ready's checks map to the specifications below. Each entry links to the ca
 ## Accessibility
-- **WCAG 2.2 + layout stability** \u2014 The accessibility tree \u2014 image text alternatives, form labels, control names, resolved ARIA references, named iframes, a clean heading outline \u2014 is what assistive tech and AI agents parse to act on a page; explicit media dimensions and enabled zoom keep it stable and usable (with a static CLS proxy). Scored as a separate accessibilityScore, not part of the Vercel score. Canonical: <https://www.w3.org/TR/WCAG22/> Checks: A1\u2013A9.
-`;
+- **WCAG 2.2 + layout stability** \u2014 The accessibility tree \u2014 image text alternatives, form labels, control names, resolved ARIA references, named iframes, a page title, a valid language tag, a main landmark, accessible names that match their visible text, valid ARIA, well-formed lists and tables, and autofill tokens an agent can act on \u2014 is what assistive tech and AI agents parse to act on a page; explicit media dimensions and enabled zoom keep it stable and usable (with a static CLS proxy). Scored as a separate accessibilityScore, not part of the Vercel score. Canonical: <https://www.w3.org/TR/WCAG22/> Checks: A1\u2013A23.`;
 // src/resources.ts
 function registerResources(server) {
@@ -31844,7 +31860,7 @@ var validateStructuredDataInputShape = {
 // src/server.ts
 var SERVER_INFO = {
   name: "agent-ready",
-  version: "0.7.0"
+  version: "0.7.2"
 };
 function createMcpServer(config2) {
   const server = new McpServer(SERVER_INFO);

package/package.json CHANGED Viewed

@@ -1,6 +1,6 @@
 {
   "name": "agent-ready-mcp",
-  "version": "0.7.0",
+  "version": "0.7.2",
   "mcpName": "io.github.mlava/agent-ready-mcp",
   "description": "MCP server for Agent Ready — scan any URL for AI-readability against the Vercel Agent Readability Spec, the llmstxt.org standard, and agent-protocol manifests (MCP server cards, A2A, agents.json, agent-permissions.json, UCP, x402, NLWeb). 69 checks with per-check fix guidance.",
   "license": "MIT",