emorahealth.com WebMCP audit

Name: emorahealth.com WebMCP audit
Creator: ora

Rescan

audited Aug 28, 2026in 10s

47/ 100

10 tools registered. Strongest in WebMCP use, weakest in human experience.

WebMCP use59

Usefulness59

Human experience0

about:blank

No agent run recorded yet.

No steps recorded

Tools

start_here

[Step 1 of find_a_clinician · optional] Best first action for a user describing a concern. Runs a parallel lookup across crisis screening, provider availability, and the article corpus, then returns the recommended path (crisis | evaluation | self-help | mixed) with concrete next steps. Optimized for the agent's first turn — a single call replaces 2-3 sequential lookups. Use when: The user has just described a concern and you don't yet know if they need crisis routing, a clinician, articles, or all three. Call this FIRST before find_provider / search_content. Don't use when: You already have a chosen provider and just need availability or a booking URL — use check_availability or book_appointment instead. Example: start_here({ concern: "My 12-year-old has trouble sleeping and seems anxious", state: "Florida", age: 12 })Answers a question. Declared read only.

pagehttps://www.emorahealth.com/

implementation

viaimperative

entry pointnavigator

registered after1821ms

executepresent

api surface

concernrequired

ageoptional

stateoptional

insuranceoptional

annotations

read onlytrue

destructivenot declared

idempotentnot declared

open worldnot declared

untrusted contenttrue

titlenot declared

4 findings

failCanonical entry point

0 / 4

Every tool registers through the deprecated navigator.modelContext alias; move them to document.modelContext before the alias is removed.

fix

// Use the canonical entry point - navigator.modelContext is deprecated.
document.modelContext.provideContext({ tools: [/* ... */] });

warningSchema quality

0.4 / 4

9 of 10 tool schemas are harder for an agent to use than they need to be: start_here description is 835 characters, over Chrome's 500-character guidance.

start_here: description is 835 characters, over Chrome's 500-character guidance; find_provider: description is 1083 characters, over Chrome's 500-character guidance; find_provider: 4 parameters have n

fix

{
  name: "start_here",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}

warningAnnotation mismatch

2.1 / 3

3 tools declare readOnlyHint: true although the name says the call writes (start_here).

start_here: readOnlyHint=true; book_matching_session: readOnlyHint=true; book_appointment: readOnlyHint=true

fix

// Set destructiveHint/idempotentHint to match what this tool actually does.
{ name: "start_here", annotations: { destructiveHint: true } }

warningHuman parity

0 / 8

10 of 10 tools run through the JavaScript API with no visible counterpart, so the person co-browsing cannot see what the agent is offered. Declarative forms (or at least a title annotation) close the gap.

start_here, find_provider, check_availability, about_emora, search_content

fix

<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="start_here" tooldescription="...">
  <!-- the same action, as UI -->
</form>

tool json

{
  "name": "start_here",
  "description": "[Step 1 of find_a_clinician · optional]\n\nBest first action for a user describing a concern. Runs a parallel lookup across crisis screening, provider availability, and the article corpus, then returns the recommended path (crisis | evaluation | self-help | mixed) with concrete next steps. Optimized for the agent's first turn — a single call replaces 2-3 sequential lookups.\n\nUse when: The user has just described a concern and you don't yet know if they need crisis routing, a clinician, articles, or all three. Call this FIRST before find_provider / search_content.\n\nDon't use when: You already have a chosen provider and just need availability or a booking URL — use check_availability or book_appointment instead.\n\nExample: start_here({ concern: \"My 12-year-old has trouble sleeping and seems anxious\", state: \"Florida\", age: 12 })",
  "inputSchema": {
    "type": "object",
    "required": [
      "concern"
    ],
    "properties": {
      "concern": {
        "type": "string",
        "description": "Free-text description of what the user is experiencing or concerned about."
      },
      "age": {
        "type": "number",
        "description": "Client age in years (0–25+)."
      },
      "state": {
        "type": "string",
        "description": "U.S. state where the client resides.",
        "enum": [
          "Florida",
          "Texas",
          "Illinois",
          "North Carolina",
          "Georgia",
          "Ohio",
          "Missouri",
          "Arizona",
          "Wisconsin",
          "South Carolina",
          "Indiana"
        ]
      },
      "insurance": {
        "type": "string",
        "description": "Insurance plan name.",
        "enum": [
          "Cash Pay",
          "Aetna",
          "Anthem Blue Cross",
          "Baylor Scott & White",
          "Blue Cross Blue Shield",
          "ChampVA",
          "Cigna",
          "Florida Blue",
          "Magellan",
          "Medical Mutual",
          "United Healthcare",
          "Optum",
          "Oscar",
          "Tricare",
          "Tricare West",
          "UMR"
        ]
      }
    }
  },
  "annotations": {
    "readOnlyHint": true,
    "untrustedContentHint": true
  }
}

Findings

WebMCP use

59 / 100 · weight 50

passTools registered

4 / 4

10 tools registered across 1 page.

failReal-browser eligible

0 / 8

The page's code registers these tools, but a real visitor's browser does not get them yet: WebMCP needs a native modelContext or a current origin trial token, and this page has neither.

warningRegistration timing

2 / 4

The slowest tool took 1821.7000000476837ms to register, past the 1000ms an agent reading the registry at first paint would wait for.

toolget_cost_estimate

get_cost_estimate: registered at 1821.7000000476837ms

fix

// Register tools as soon as they're ready, not behind a deferred/async chunk.
document.modelContext.provideContext({ tools: [/* ... */] });

failCanonical entry point

0 / 4

Every tool registers through the deprecated navigator.modelContext alias; move them to document.modelContext before the alias is removed.

toolstart_here

fix

// Use the canonical entry point - navigator.modelContext is deprecated.
document.modelContext.provideContext({ tools: [/* ... */] });

passSchema validity

5 / 5

All 10 tools declare a structurally valid object input schema.

warningSchema quality

0.4 / 4

9 of 10 tool schemas are harder for an agent to use than they need to be: start_here description is 835 characters, over Chrome's 500-character guidance.

toolstart_here

start_here: description is 835 characters, over Chrome's 500-character guidance; find_provider: description is 1083 characters, over Chrome's 500-character guidance; find_provider: 4 parameters have n

fix

{
  name: "start_here",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}

passStub detection

5 / 5

All 10 tools declare an execute handler.

passAnnotations present

4 / 4

All 10 tools declare a readOnlyHint, so an agent knows which calls change state.

warningAnnotation mismatch

2.1 / 3

3 tools declare readOnlyHint: true although the name says the call writes (start_here).

toolstart_here

start_here: readOnlyHint=true; book_matching_session: readOnlyHint=true; book_appointment: readOnlyHint=true

fix

// Set destructiveHint/idempotentHint to match what this tool actually does.
{ name: "start_here", annotations: { destructiveHint: true } }

passUntrusted content hint

2 / 2

All 1 tool that looks like it returns third-party text declares untrustedContentHint.

passRegistration errors

2 / 2

No tool registration threw during the capture.

warningToolchange coherence

1 / 2

10 tools were registered but the page never dispatched a toolchange event, so an agent subscribed to registry updates never learns they exist.

fix

document.modelContext.dispatchEvent(new Event("toolchange"));

not applicableInjection surface

0 / 3

The injection-surface model hop returned no usable result, so tool contracts were not reviewed for injection surface.

Usefulness

59 / 100 · weight 30

warningNaming quality

1.9 / 2

Tool naming makes selection harder than it needs to be: 1 name does not start with a verb (about_emora).

toolabout_emora

fix

{ name: "about_emora" /* short, unique, verb-based */ }

not applicableDescription quality

0 / 6

The description-quality model hop returned no usable result, so descriptions were not rated.

not applicableTool selection

0 / 14

The tool-selection model hop returned no usable result.

warningCoverage vs. site type

4 / 8

The tool set covers 2 of the 4 things an agent needs on a commerce site; it cannot add to or read the cart or start checkout or place an order.

Human experience

0 / 100 · weight 20

warningHuman parity

0 / 8

toolstart_here

start_here, find_provider, check_availability, about_emora, search_content

fix

<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="start_here" tooldescription="...">
  <!-- the same action, as UI -->
</form>

not applicablePage experience

0 / 12

The page-experience model hop returned no usable result, so the page was not graded.

Tool selection

The tool-selection model hop returned no usable result.

how this was captured

chrome: 148.0.7778.96
capture shim: v1
spec snapshot: 2026-08
mode: fast
pages: 1
tools: 10
run time: 10s
finished: Aug 28, 2026