Clinical trials across multiple registries with identifiers and source links.
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.
Fetch one canonical trial by id, with every registry's provenance + cross-registry IDs.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http trialbase https://api.trialbase-db.org/mcp
Fetch one canonical trial by id, with every registry's provenance + cross-registry IDs.
| Argument |
|---|
| Type |
|---|
| Description |
|---|
| id* | integer |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"id"
],
"properties": {
"id": {
"type": "integer",
"maximum": 9007199254740991,
"exclusiveMinimum": 0
}
}
}Search trials by condition, intervention, sponsor, phase, status, country, or free text.
| Argument | Type | Description |
|---|---|---|
| q | string | |
| limit | integer | |
| phase | string | |
| cursor | string | |
| status | string | |
| country | string | |
| sponsor | string | |
| registry | string | |
| condition | string | |
| study_type | string | |
| intervention | string |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {
"q": {
"type": "string",
"maxLength": 160
},
"limit": {
"type": "integer",
"maximum": 100,
"minimum": 1
},
"phase": {
"type": "string",
"maxLength": 20
},
"cursor": {
"type": "string"
},
"status": {
"type": "string",
"maxLength": 40
},
"country": {
"type": "string",
"maxLength": 80
},
"sponsor": {
"type": "string",
"maxLength": 160
},
"registry": {
"type": "string",
"maxLength": 40
},
"condition": {
"type": "string",
"maxLength": 160
},
"study_type": {
"type": "string",
"maxLength": 20
},
"intervention": {
"type": "string",
"maxLength": 160
}
}
}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 12, 2026.