MCP Tools
Integrate Midpage's caselaw database with AI assistants using the Model Context Protocol.
Setup
Claude Desktop
Add this to your Claude Desktop configuration:
{
"mcpServers": {
"midpage": {
"url": "https://api.midpage.ai/mcp"
}
}
}
Cursor
In Cursor settings, add the MCP server:
https://api.midpage.ai/mcp
Available Tools
searchAndReview
Search for cases and evaluate them against a specific legal question.
Parameters:
| Parameter | Type | Description |
|---|---|---|
searches |
array | Array of search queries |
searches[].query |
string | Legal proposition or text to search |
searches[].jurisdictionType |
string | "state", "federal", or "state_and_federal" |
searches[].circuits |
string[] | Federal circuits (e.g., ["9", "2"]) |
searches[].states |
string[] | State names (e.g., ["California", "New York"]) |
searches[].startDate |
string | Start date (YYYY-MM-DD) |
searches[].endDate |
string | End date (YYYY-MM-DD) |
prompt |
string | Question to evaluate each case against |
Example:
{
"searches": [{
"query": "breach of fiduciary duty elements",
"jurisdictionType": "federal",
"circuits": ["9"]
}],
"prompt": "Does this case establish the elements required to prove breach of fiduciary duty?"
}
reviewDocuments
Review specific cases by citation or opinion ID.
Parameters:
| Parameter | Type | Description |
|---|---|---|
reporterCitations |
string[] | Citations like "556 U.S. 662" |
opinionIds |
string[] | Numeric opinion IDs |
dockets |
array | Court and docket number pairs |
prompt |
string | Question to evaluate each document against |
Example:
{
"reporterCitations": ["556 U.S. 662", "550 U.S. 544"],
"prompt": "What standard does this case establish for pleading requirements?"
}
researchAgent
Conduct comprehensive legal research on a topic.
Parameters:
| Parameter | Type | Description |
|---|---|---|
researchGoal |
string | Clear, specific research goal |
Example:
{
"researchGoal": "Find Ninth Circuit cases on whether property boundary mistakes satisfy adverse possession 'hostility' requirement"
}