developers.openai.com WebMCP audit

Rescan
audited Aug 28, 2026in 4s
66/ 100

5 tools registered. Strongest in usefulness, weakest in human experience.

WebMCP use73
Usefulness98
Human experience0
about:blank

No agent run recorded yet.

No steps recorded

Tools

/
search_openai_docs: warning

search_openai_docs

Search OpenAI Developers and ChatGPT Learn documentation. Use lookup_page on a result to read the exact page content.Answers a question. Declared read only.
pagehttps://developers.openai.com/
implementation
viaimperative
entry pointdocument
registered after578ms
executepresent
api surface
queryrequired
limitoptional
cursoroptional
annotations
read onlytrue
destructivenot declared
idempotentnot declared
open worldnot declared
untrusted contenttrue
titlenot declared
2 findings
warningSchema quality
0.8 / 4
4 of 5 tool schemas are harder for an agent to use than they need to be: search_openai_docs 3 parameters have no description (query, limit, cursor).
search_openai_docs: 3 parameters have no description (query, limit, cursor); lookup_page: 2 parameters have no description (path, anchor); navigate_to_page: 1 parameter has no description (intent); ge
fix
{
  name: "search_openai_docs",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}
warningHuman parity
0 / 8
5 of 5 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.
search_openai_docs, lookup_page, lookup_context, navigate_to_page, generate_custom_guide
fix
<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="search_openai_docs" tooldescription="...">
  <!-- the same action, as UI -->
</form>
tool json
{
  "name": "search_openai_docs",
  "description": "Search OpenAI Developers and ChatGPT Learn documentation. Use lookup_page on a result to read the exact page content.",
  "inputSchema": {
    "type": "object",
    "additionalProperties": false,
    "properties": {
      "query": {
        "type": "string",
        "minLength": 1,
        "maxLength": 500
      },
      "limit": {
        "type": "integer",
        "minimum": 1,
        "maximum": 50
      },
      "cursor": {
        "type": "string",
        "minLength": 1,
        "maxLength": 2000
      }
    },
    "required": [
      "query"
    ]
  },
  "annotations": {
    "readOnlyHint": true,
    "untrustedContentHint": true
  }
}
Showing search_openai_docs

Findings

WebMCP use

73 / 100 · weight 50
passTools registered
4 / 4
5 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.
passRegistration timing
4 / 4
Every measured tool registered within 577.7999999523163ms of navigation.
toollookup_page
passCanonical entry point
4 / 4
All 5 tools register on the canonical document.modelContext entry point.
passSchema validity
5 / 5
All 5 tools declare a structurally valid object input schema.
warningSchema quality
0.8 / 4
4 of 5 tool schemas are harder for an agent to use than they need to be: search_openai_docs 3 parameters have no description (query, limit, cursor).
toolsearch_openai_docs
search_openai_docs: 3 parameters have no description (query, limit, cursor); lookup_page: 2 parameters have no description (path, anchor); navigate_to_page: 1 parameter has no description (intent); ge
fix
{
  name: "search_openai_docs",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}
passStub detection
5 / 5
All 5 tools declare an execute handler.
passAnnotations present
4 / 4
All 5 tools declare a readOnlyHint, so an agent knows which calls change state.
passAnnotation mismatch
3 / 3
No tool claims to be read-only while its name says it writes (5 declared hints checked).
not applicableUntrusted content hint
0 / 2
No tool's contract suggests it returns text written by somebody other than the site, so there is nothing to flag as untrusted.
passRegistration errors
2 / 2
No tool registration threw during the capture.
warningToolchange coherence
1 / 2
5 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

98 / 100 · weight 30
warningNaming quality
1.8 / 2
Tool naming makes selection harder than it needs to be: 1 name does not start with a verb (navigate_to_page).
toolnavigate_to_page
fix
{ name: "navigate_to_page" /* 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.
passCoverage vs. site type
8 / 8
The tool set covers the whole core docs flow: search or browse the catalog, read a page or article, navigate the documentation tree.

Human experience

0 / 100 · weight 20
warningHuman parity
0 / 8
5 of 5 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.
toolsearch_openai_docs
search_openai_docs, lookup_page, lookup_context, navigate_to_page, generate_custom_guide
fix
<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="search_openai_docs" 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
5
run time
4s
finished
Aug 28, 2026