Public AI claim register: E0-E4 evidence grades, forecast tracking, no verdicts.
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.
List register forecasts whose due date falls within the given number of days (including those already due and still open). Futures essays are not included. Signetex grades evidence and issues no true/false verdicts.
Include forecasts due on or before today plus this many days (default 365, max 3650). Already-due open forecasts are included.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
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://mcp.signetex.ch/mcp
List register forecasts whose due date falls within the given number of days (including those already due and still open). Futures essays are not included. Signetex grades evidence and issues no true/false verdicts.
| Argument | Type | Description |
|---|---|---|
| within_days | integer | Include forecasts due on or before today plus this many days (default 365, max 3650). Already-due open forecasts are included. |
{
"type": "object",
"properties": {
"within_days": {
"type": "integer",
"maximum": 3650,
"minimum": 0,
"description": "Include forecasts due on or before today plus this many days (default 365, max 3650). Already-due open forecasts are included."
}
},
"additionalProperties": false
}Return the published E0–E4 evidence-grade definitions from the Signetex method page, not a true/false scale.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Get one register entry by its stable ID (for example SGX-AI-000031 or SGX-HU-000016), including sources, counter-evidence, grade rationale and history, with the canonical page URL. Signetex grades evidence and issues no true/false verdicts.
| Argument | Type | Description |
|---|---|---|
| id* | string | Stable Signetex entry ID, for example SGX-AI-000031. |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Stable Signetex entry ID, for example SGX-AI-000031."
}
},
"additionalProperties": false
}Return the latest published OpenTimestamps proof snapshot of the register (root hash, status, manifest URLs).
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Headline counts for the published register: totals, areas, grades, forecast outcomes. A hit rate is only included once at least 30 forecasts have closed, as on the public site. Signetex grades evidence and issues no true/false verdicts.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}Search the Signetex register of claims about AI consciousness, AGI timelines, frontier capabilities, and AI and humans. Returns matching entries with canonical URLs. Does not include the speculative Futures essays. Signetex grades evidence and issues no true/false verdicts.
| Argument | Type | Description |
|---|---|---|
| area | string | Register area: machine-intelligence (the AI itself) or human-impact (AI and humans). |
| kind | string | Entry kind: finding, forecast or possibility (also accepts the stored values befund, prognose, moeglichkeit). |
| grade | string | Evidence grade: E0, E1, E2, E3 or E4. |
| limit | integer | Maximum number of results to return (default 10, max 50). |
| query | string | Free-text search over English and German claim text, speaker, organisation, topic and ID. Optional if filters are set. |
| outcome | string | Forecast outcome: open, met, partial, not_met, unclear (German stored values also accepted). |
{
"type": "object",
"properties": {
"area": {
"enum": [
"machine-intelligence",
"human-impact"
],
"type": "string",
"description": "Register area: machine-intelligence (the AI itself) or human-impact (AI and humans)."
},
"kind": {
"type": "string",
"description": "Entry kind: finding, forecast or possibility (also accepts the stored values befund, prognose, moeglichkeit)."
},
"grade": {
"type": "string",
"description": "Evidence grade: E0, E1, E2, E3 or E4."
},
"limit": {
"type": "integer",
"maximum": 50,
"minimum": 1,
"description": "Maximum number of results to return (default 10, max 50)."
},
"query": {
"type": "string",
"description": "Free-text search over English and German claim text, speaker, organisation, topic and ID. Optional if filters are set."
},
"outcome": {
"type": "string",
"description": "Forecast outcome: open, met, partial, not_met, unclear (German stored values also accepted)."
}
},
"additionalProperties": false
}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 17, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.