Register of AI claims in mathematics, graded by the evidence behind each. Paid Lean proof checks.
Maintain this server? Claim it
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.
Benchmarks as sets of statements, with how many members already carry public AI work: a contamination figure rather than a study.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http register https://qed.bot/mcp
Benchmarks as sets of statements, with how many members already carry public AI work: a contamination figure rather than a study.
| Argument | Type | Description |
|---|---|---|
| include_members | boolean |
{
"type": "object",
"properties": {
"include_members": {
"type": "boolean",
"default": false
}
}
}Have qed.bot rebuild a Lean file at its commit and read what every theorem in it depends on, under the admission rules every stranger's file is held to. Costs one credit, or two for a private check; buy credits and make an API key at https://qed.bot/check, and send it as Authorization: Bearer qed_.... Returns at once with the check queued; call get_check until it is done, which takes minutes. A file already checked in public is returned free, and a check that stops on qed.bot's side is refunded.
| Argument | Type | Description |
|---|---|---|
| url* | string | GitHub blob URL of the .lean file, at a full commit |
| private | boolean | Only you can see the result (2 credits) |
| theorem | string | The theorem it proves, which must be declared |
{
"type": "object",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"description": "GitHub blob URL of the .lean file, at a full commit"
},
"private": {
"type": "boolean",
"default": false,
"description": "Only you can see the result (2 credits)"
},
"theorem": {
"type": "string",
"description": "The theorem it proves, which must be declared"
}
}
}Results recorded by more than one registry, and how the positions differ. A disagreement is a difference of judgement about who did the mathematics; a gap is one registry holding stronger evidence than another; a lead is a proof somebody else records that this register has not rebuilt.
| Argument | Type | Description |
|---|---|---|
| limit | integer | |
| finding | string | |
| registry | string | Restrict to results this registry holds |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 25
},
"finding": {
"enum": [
"any",
"disagreement",
"gap",
"lead",
"agreement"
],
"type": "string",
"default": "any"
},
"registry": {
"type": "string",
"description": "Restrict to results this registry holds"
}
}
}Announcements of results, graded for how much of what they claim can be inspected: whether each result is named, written up, and published with a checkable artifact.
| Argument | Type | Description |
|---|---|---|
| org | string | Restrict to one organisation |
{
"type": "object",
"properties": {
"org": {
"type": "string",
"description": "Restrict to one organisation"
}
}
}A paid check's status and verdict, by the id check_proof returned. Needs the same API key for a private check.
| Argument | Type | Description |
|---|---|---|
| id* | integer | The check's id |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "integer",
"description": "The check's id"
}
}
}One declaration in full: its sources, related formalisations, main results, the results held here it joins and on what key, Palomar's registrations of it, and any contradiction between what it declares and what a rebuild found.
| Argument | Type | Description |
|---|---|---|
| id* | string | owner/repo for a root declaration, owner/repo/path, or its GitHub URL |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "owner/repo for a root declaration, owner/repo/path, or its GitHub URL"
}
}
}Everything the register holds about one statement: formalisations, cited proofs, recompilation results, graded claims, relations and aliases.
| Argument | Type | Description |
|---|---|---|
| id* | string | Canonical id or any alias, e.g. erdos:707 |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Canonical id or any alias, e.g. erdos:707"
}
}
}Everything the register holds about one target: the record history with who held each value and when, the graded claims, and any check run here.
| Argument | Type | Description |
|---|---|---|
| id* | string | Canonical id, e.g. capset:dimension-8 |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Canonical id, e.g. capset:dimension-8"
}
}
}The collections held, with counts of statements, formalisations and AI activity.
No arguments.
{
"type": "object",
"properties": {}
}The object checkers available, what each establishes, what an artifact must look like, and which targets have been checked with it. Read this before calling submit_construction, which only the local server offers.
No arguments.
{
"type": "object",
"properties": {}
}Track record per AI system, computed from recorded claims. Not a benchmark: which statements attract attention has little to do with how hard they are.
| Argument | Type | Description |
|---|---|---|
| name | string | Restrict to one system |
| limit | integer |
{
"type": "object",
"properties": {
"name": {
"type": "string",
"description": "Restrict to one system"
},
"limit": {
"type": "integer",
"default": 25
}
}
}Statements not recorded as solved, most valuable prize first. Use formalised_only to restrict to those a prover could attempt directly.
| Argument | Type | Description |
|---|---|---|
| limit | integer | |
| min_prize | integer | |
| collection | string | |
| formalised_only | boolean |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 25
},
"min_prize": {
"type": "integer",
"default": 0
},
"collection": {
"type": "string"
},
"formalised_only": {
"type": "boolean"
}
}
}Every registry's position on one result, this register included, with each one's own label preserved beside the bucket it was placed in.
| Argument | Type | Description |
|---|---|---|
| id* | string | Node id, e.g. erdos:346 |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Node id, e.g. erdos:346"
}
}
}Every formalization.yaml the register has read, with what each declares about its own axioms, unproved goals, assumed literature, automation and review, and who has checked it: this register, Palomar, or nobody independent of its authors.
| Argument | Type | Description |
|---|---|---|
| flag | string | |
| limit | integer | |
| query | string | Text to match in name, repository, path, authors or models |
| method | string | |
| review | string | |
| held_only | boolean | Only declarations joined to a result held here |
| checked_by | string | Who checked it; none for declarations nobody independent has checked |
{
"type": "object",
"properties": {
"flag": {
"enum": [
"alignment",
"axioms",
"comparator",
"contradicted",
"divergences",
"literature",
"original",
"unproved"
],
"type": "string"
},
"limit": {
"type": "integer",
"default": 25
},
"query": {
"type": "string",
"description": "Text to match in name, repository, path, authors or models"
},
"method": {
"enum": [
"autonomous",
"agent",
"copilot",
"manual",
"other"
],
"type": "string"
},
"review": {
"enum": [
"peer-reviewed",
"author-verified",
"agent-reviewed",
"self-assessed",
"unchecked",
"other",
"not stated"
],
"type": "string"
},
"held_only": {
"type": "boolean",
"description": "Only declarations joined to a result held here"
},
"checked_by": {
"enum": [
"Palomar",
"qed.bot",
"none"
],
"type": "string",
"description": "Who checked it; none for declarations nobody independent has checked"
}
}
}Search the register of mathematical statements. Returns summaries with formal status, prizes, and how much AI work is recorded against each.
| Argument | Type | Description |
|---|---|---|
| limit | integer | |
| query | string | Text to match in id, title, collection or tags |
| has_prize | boolean | |
| collection | string | Restrict to one collection |
| has_ai_claim | boolean | |
| formal_status | string | |
| verified_only | boolean | Only statements with a proof recompiled here |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 25
},
"query": {
"type": "string",
"description": "Text to match in id, title, collection or tags"
},
"has_prize": {
"type": "boolean"
},
"collection": {
"type": "string",
"description": "Restrict to one collection"
},
"has_ai_claim": {
"type": "boolean"
},
"formal_status": {
"enum": [
"open",
"solved",
"mixed",
"unclassified"
],
"type": "string"
},
"verified_only": {
"type": "boolean",
"description": "Only statements with a proof recompiled here"
}
}
}Search the targets: quantities with a direction and a record history, where AI work moved a bound or produced a better object rather than settling a statement.
| Argument | Type | Description |
|---|---|---|
| limit | integer | |
| query | string | Text to match in id, name, collection or tags |
| standing | string | |
| collection | string | |
| checked_only | boolean | Only targets whose object was rebuilt here |
| held_by_machine | boolean | Only where the leading record is a machine's |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 25
},
"query": {
"type": "string",
"description": "Text to match in id, name, collection or tags"
},
"standing": {
"enum": [
"record",
"matched",
"superseded",
"below",
"unstated"
],
"type": "string"
},
"collection": {
"type": "string"
},
"checked_only": {
"type": "boolean",
"description": "Only targets whose object was rebuilt here"
},
"held_by_machine": {
"type": "boolean",
"description": "Only where the leading record is a machine's"
}
}
}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 21, 2026.