Claim check for agents: supported, refuted or unverifiable with sources. Canada Site Ops.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Bearer <GroundLock API key> from https://sage-cedar-nova-otter.grok.me/signup
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.
Check a factual claim against public pages via GroundLock. Returns verdict (supported|refuted|unverifiable), confidence 0-1, and sources with quoted snippets. Not a guarantee.
Factual claim to check (1-2000 chars)
Work budget: quick 1 page, standard up to 5, deep up to 10. Default quick.
GroundLock API key if not sent as Authorization Bearer. Free at https://sage-cedar-nova-otter.grok.me/signup
Optional context (up to 4000 chars)
Up to 8 public URLs to read first
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 groundlock-mcp https://groundlock-mcp.marc-cf2.workers.dev/mcp \ --header 'Authorization: Bearer YOUR_TOKEN'
Headers
Authorization (required): Bearer <GroundLock API key> from https://sage-cedar-nova-otter.grok.me/signupCheck a factual claim against public pages via GroundLock. Returns verdict (supported|refuted|unverifiable), confidence 0-1, and sources with quoted snippets. Not a guarantee.
| Argument | Type | Description |
|---|
| claim* | string | Factual claim to check (1-2000 chars) |
| depth | string | Work budget: quick 1 page, standard up to 5, deep up to 10. Default quick. |
| api_key | string | GroundLock API key if not sent as Authorization Bearer. Free at https://sage-cedar-nova-otter.grok.me/signup |
| context | string | Optional context (up to 4000 chars) |
| sources | array | Up to 8 public URLs to read first |
{
"type": "object",
"required": [
"claim"
],
"properties": {
"claim": {
"type": "string",
"description": "Factual claim to check (1-2000 chars)"
},
"depth": {
"enum": [
"quick",
"standard",
"deep"
],
"type": "string",
"description": "Work budget: quick 1 page, standard up to 5, deep up to 10. Default quick."
},
"api_key": {
"type": "string",
"description": "GroundLock API key if not sent as Authorization Bearer. Free at https://sage-cedar-nova-otter.grok.me/signup"
},
"context": {
"type": "string",
"description": "Optional context (up to 4000 chars)"
},
"sources": {
"type": "array",
"items": {
"type": "string"
},
"description": "Up to 8 public URLs to read first"
}
}
}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 6, 2026.
Takes a password, API key or token as an argument, so the secret passes through the model's context and chat logs.
Fix: Read credentials from the server's own configuration (environment, OAuth) instead of a tool argument.
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.