Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Two-proportion z test for an A/B test: the two rates, the lift, the p-value, a 95% range for the difference, and visitors needed per version to detect this gap.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http monte https://mcp.monte.markets/mcp
Two-proportion z test for an A/B test: the two rates, the lift, the p-value, a 95% range for the difference, and visitors needed per version to detect this gap.
| Argument | Type | Description |
|---|---|---|
| visitors_a* | integer | |
| visitors_b* | integer | |
| conversions_a* | integer | |
| conversions_b* | integer |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"visitors_a",
"conversions_a",
"visitors_b",
"conversions_b"
],
"properties": {
"visitors_a": {
"type": "integer",
"maximum": 9007199254740991,
"exclusiveMinimum": 0
},
"visitors_b": {
"type": "integer",
"maximum": 9007199254740991,
"exclusiveMinimum": 0
},
"conversions_a": {
"type": "integer",
"maximum": 9007199254740991,
"minimum": 0
},
"conversions_b": {
"type": "integer",
"maximum": 9007199254740991,
"minimum": 0
}
}
}One topic in plain English: the idea, why it matters, a worked example, the usual trap, the formulas and the five lessons, with links.
| Argument | Type | Description |
|---|---|---|
| topic* | string | Topic id from list_topics, like normal or clt |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"topic"
],
"properties": {
"topic": {
"enum": [
"data",
"center",
"spread",
"prob",
"rv",
"two",
"count",
"poisson",
"normal",
"clt",
"ci",
"test",
"reg",
"mreg"
],
"type": "string",
"description": "Topic id from list_topics, like normal or clt"
}
}
}The 11 solvers and the inputs each one takes (key, label, default value, and choices where there are some). Use with solve.
No arguments.
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {}
}The 14 statistics topics on Monte, in order, with their page links.
No arguments.
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {}
}Statistics terms in plain English with an example each, matched on the term or its meaning.
| Argument | Type | Description |
|---|---|---|
| query* | string | A word or phrase, like standard error or p-value |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"query"
],
"properties": {
"query": {
"type": "string",
"minLength": 1,
"description": "A word or phrase, like standard error or p-value"
}
}
}Runs one of Monte's solvers and returns every answer with its working, exactly as the solver page shows it. Inputs you leave out keep the solver's defaults. Numbers may be written as 0.7, 70% or 7/10.
| Argument | Type | Description |
|---|---|---|
| inputs | object | Input key to value, as text, for example {"mu": "500", "sd": "20"} |
| solver* | string | Solver id from list_solvers, like tree, normal or ci |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"solver"
],
"properties": {
"inputs": {
"type": "object",
"description": "Input key to value, as text, for example {\"mu\": \"500\", \"sd\": \"20\"}",
"propertyNames": {
"type": "string"
},
"additionalProperties": {
"type": "string"
}
},
"solver": {
"enum": [
"tree",
"draws",
"events",
"trials",
"poisson",
"normal",
"samp",
"ci",
"test",
"bday",
"data"
],
"type": "string",
"description": "Solver id from list_solvers, like tree, normal or ci"
}
}
}For a positive result, the chance you really have the thing, from how common it is, how often the test catches it and how often it false-alarms (Bayes).
| Argument | Type | Description |
|---|---|---|
| catches* | number | Sensitivity: chance the test is positive when you have it, 0 to 1 |
| how_common* | number | Base rate, 0 to 1, like 0.01 for 1 in 100 |
| false_alarms* | number | Chance the test is positive when you do not have it, 0 to 1 |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"how_common",
"catches",
"false_alarms"
],
"properties": {
"catches": {
"type": "number",
"maximum": 1,
"minimum": 0,
"description": "Sensitivity: chance the test is positive when you have it, 0 to 1"
},
"how_common": {
"type": "number",
"maximum": 1,
"minimum": 0,
"description": "Base rate, 0 to 1, like 0.01 for 1 in 100"
},
"false_alarms": {
"type": "number",
"maximum": 1,
"minimum": 0,
"description": "Chance the test is positive when you do not have it, 0 to 1"
}
}
}Expected cost of buying versus skipping a warranty or insurance plan, the break-even chance and the plan's markup.
| Argument | Type | Description |
|---|---|---|
| cost* | number | What it costs you if it happens and you have no plan, in dollars |
| price* | number | Price of the plan, in dollars |
| chance* | number | Chance the bad thing happens over the whole plan, 0 to 1 |
| deductible | number | What you still pay with the plan, in dollars |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"price",
"chance",
"cost"
],
"properties": {
"cost": {
"type": "number",
"description": "What it costs you if it happens and you have no plan, in dollars",
"exclusiveMinimum": 0
},
"price": {
"type": "number",
"minimum": 0,
"description": "Price of the plan, in dollars"
},
"chance": {
"type": "number",
"maximum": 1,
"minimum": 0,
"description": "Chance the bad thing happens over the whole plan, 0 to 1"
},
"deductible": {
"type": "number",
"minimum": 0,
"description": "What you still pay with the plan, in dollars"
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 19, 2026.