tesla-daily.lugassy.net WebMCP audit

Rescan
audited Aug 28, 2026in 3s
59/ 100

2 tools registered. Strongest in usefulness, weakest in WebMCP use.

WebMCP use38
Usefulness100
Human experience50
about:blank

No agent run recorded yet.

No steps recorded

Tools

/
search_episodes: fail

search_episodes

Search this podcast by topic, person, or keyword. Returns ranked results with title, date, URL, and a transcript snippet.Registered without an execute function, so calling it does nothing.
pagehttps://tesla-daily.lugassy.net/
implementation
viaimperative
entry pointnavigator
registered after53ms
executemissing
api surface
queryrequired
limitoptional
annotations
read onlynot declared
destructivenot declared
idempotentnot declared
open worldnot declared
untrusted contentnot declared
titlenot declared
6 findings
passRegistration timing
4 / 4
Every measured tool registered within 53.19999980926514ms of navigation. 1 further tool carried no measurable registration time.
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
2 / 4
1 of 2 tool schemas are harder for an agent to use than they need to be: search_episodes 1 parameter has no description (limit).
search_episodes: 1 parameter has no description (limit)
fix
{
  name: "search_episodes",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}
failStub detection
0 / 5
Every registered tool is a facade: none declares an execute handler, so an agent that calls one gets nothing back.
search_episodes
fix
{
  name: "search_episodes",
  execute: async (args) => {
    // Perform the real action here - return live data, not a static placeholder.
    return result;
  }
}
failAnnotations present
0 / 4
1 of 1 tools declare no readOnlyHint, so an agent cannot tell a read from a write without guessing from the name.
search_episodes
fix
{
  name: "search_episodes",
  annotations: { readOnlyHint: true, destructiveHint: false, idempotentHint: true }
}
warningHuman parity
4 / 8
1 of 2 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_episodes
fix
<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="search_episodes" tooldescription="...">
  <!-- the same action, as UI -->
</form>
tool json
{
  "name": "search_episodes",
  "description": "Search this podcast by topic, person, or keyword. Returns ranked results with title, date, URL, and a transcript snippet.",
  "inputSchema": {
    "type": "object",
    "properties": {
      "query": {
        "type": "string",
        "description": "Search query (free text)."
      },
      "limit": {
        "type": "integer",
        "default": 10,
        "minimum": 1,
        "maximum": 50
      }
    },
    "required": [
      "query"
    ]
  },
  "annotations": {}
}
Showing search_episodes

Findings

WebMCP use

38 / 100 · weight 50
warningTools registered
2 / 4
1 imperative tool and 1 declarative form registered across 1 page, but not one of the 1 JavaScript tool declares an execute handler, so an agent finds them and can run none of them.
fix
document.modelContext.provideContext({
  tools: [{ name: "your_tool_name", description: "...", inputSchema: { type: "object", properties: {} }, execute: async (args) => { /* ... */ } }],
});
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. The 1 declarative form tool on the page stay usable regardless.
passRegistration timing
4 / 4
Every measured tool registered within 53.19999980926514ms of navigation. 1 further tool carried no measurable registration time.
toolsearch_episodes
failCanonical entry point
0 / 4
Every tool registers through the deprecated navigator.modelContext alias; move them to document.modelContext before the alias is removed.
toolsearch_episodes
fix
// Use the canonical entry point - navigator.modelContext is deprecated.
document.modelContext.provideContext({ tools: [/* ... */] });
passSchema validity
5 / 5
All 1 tool declare a structurally valid object input schema.
warningSchema quality
2 / 4
1 of 2 tool schemas are harder for an agent to use than they need to be: search_episodes 1 parameter has no description (limit).
toolsearch_episodes
search_episodes: 1 parameter has no description (limit)
fix
{
  name: "search_episodes",
  inputSchema: {
    type: "object",
    properties: {
      query: { type: "string", description: "What to search for" }
    },
    required: ["query"]
  }
}
failStub detection
0 / 5
Every registered tool is a facade: none declares an execute handler, so an agent that calls one gets nothing back.
toolsearch_episodes
search_episodes
fix
{
  name: "search_episodes",
  execute: async (args) => {
    // Perform the real action here - return live data, not a static placeholder.
    return result;
  }
}
failAnnotations present
0 / 4
1 of 1 tools declare no readOnlyHint, so an agent cannot tell a read from a write without guessing from the name.
toolsearch_episodes
search_episodes
fix
{
  name: "search_episodes",
  annotations: { readOnlyHint: true, destructiveHint: false, idempotentHint: true }
}
not applicableAnnotation mismatch
0 / 3
No tool declares a readOnlyHint, so there is no safety claim to contradict.
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
1 tool was 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

100 / 100 · weight 30
passNaming quality
2 / 2
All 2 tool names are consistent, verb-led, and within Chrome's size guidance.
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.
not applicableCoverage vs. site type
0 / 8
This tool set does not place the site in a category with a known expected tool shape, so there is no coverage baseline to score it against.

Human experience

50 / 100 · weight 20
warningHuman parity
4 / 8
1 of 2 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_episodes
search_episodes
fix
<!-- Give the tool a visible counterpart: a declarative form is one the person can see and use too. -->
<form toolname="search_episodes" 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
2
run time
3s
finished
Aug 28, 2026