MCP Integration

The Court Rules MCP server gives AI agents direct access to structured legal data: court rules for 20+ districts and 630+ judges, plus privacy enforcement actions from 21 jurisdictions (FTC, HHS, and 19 state AGs).

What you get

Seven tools across two data domains:

Domain	Tool	Purpose
Court Rules	`list_courts`	List all courts with status and judge counts
Court Rules	`search_judges`	Search judges by district, name, or type
Court Rules	`get_judge_rules`	Get all rules for a specific judge (page limits, format, procedures)
Court Rules	`check_compliance`	Generate compliance check parameters for the REST API (`/api/v1/check`)
Enforcement	`search_enforcement_actions`	Search enforcement events by jurisdiction, date, industry, violation type
Enforcement	`get_enforcement_details`	Full event details with laws cited, remedies, and source quotes
Enforcement	`get_enforcement_stats`	Summary statistics by jurisdiction, violation type, or time period

How it works

Your MCP client connects to the hosted server. Tool calls query structured court rules and enforcement data. Results include source citations and official document links.

Quick start

Add to your MCP client config (Claude Desktop, Cursor, Windsurf, or any MCP client):

{
  "mcpServers": {
    "court-rules": {
      "url": "https://mcp.courtrules.app/mcp",
      "transport": "streamable-http"
    }
  }
}

No local setup or environment variables required. Auth is handled automatically through the OAuth flow.

Try it

Open Claude Desktop and ask:

“What privacy enforcement actions happened in California this year?”

The agent will call search_enforcement_actions with jurisdiction: "CA" and return structured results with entity names, fine amounts, violation types, and source URLs.

Sample vs Full access

The hosted MCP server has two access tiers:

	Sample	Full
Auth required	No	OAuth 2.1
Enforcement events	3 recent highlights	1,100+ events
Court rules	3 courts, 3 judges	20+ courts, 630+ judges
Filters	Disabled	Full query support
Use case	Evaluate data quality and tool behavior	Production integration

Sample access returns a fixed preview of real enforcement data so you can evaluate the data shape, field quality, and tool behavior before authenticating. All sample responses include "access": "sample" and a notice explaining the limitation. To upgrade to full access, contact saurabh@courtrules.app for OAuth credentials.

Tool reference

search_enforcement_actions

Search privacy and regulatory enforcement events. Filters combine with AND logic. Parameters:

Parameter	Type	Description
`jurisdiction`	string	Filter by jurisdiction: CA, FTC, NY, TX, HHS, CT, CO, OR, NJ, VA, MN, etc.
`date_from`	string	Start date (YYYY-MM-DD)
`date_to`	string	End date (YYYY-MM-DD)
`industry`	string	Entity industry: technology, healthcare, gaming, education, financial_services, etc.
`violation_type`	string	Violation category: data_breach, opt_out_failure, children_data, security_failure, etc.
`entity_name`	string	Partial match on entity name
`risk_level`	string	critical, high, medium, or low
`limit`	number	Max results (default 20, max 100)

Example response:

{
  "results": [
    {
      "id": "evt_001",
      "title": "Attorney General Bonta Announces $2.75 Million CCPA Settlement with Disney",
      "event_date": "2026-02-11",
      "jurisdiction": "CA",
      "event_type": "settlement",
      "entity_name": "The Walt Disney Company",
      "fine_amount": 2750000,
      "violation_types": ["opt_out_failure"],
      "risk_level": "high",
      "summary": "Disney's opt-out processes failed to stop data sharing across devices. $2.75M penalty.",
      "primary_source_url": "https://oag.ca.gov/news/press-releases/..."
    }
  ],
  "total": 1
}

get_enforcement_details

Get full details for a specific enforcement event, including laws cited, remedies, and verbatim source quotes. Parameters:

Parameter	Type	Required	Description
`event_id`	string	Yes	Event UUID

Example response:

{
  "id": "evt_001",
  "title": "Attorney General Bonta Announces $2.75 Million CCPA Settlement with Disney",
  "event_date": "2026-02-11",
  "jurisdiction": "CA",
  "event_type": "settlement",
  "entity_name": "The Walt Disney Company",
  "entity_industry": "media_entertainment",
  "fine_amount": "$2,750,000",
  "consumers_affected": null,
  "violation_types": ["opt_out_failure"],
  "risk_level": "high",
  "summary": "California AG secured a $2.75M settlement after finding Disney's streaming platforms failed to honor consumer opt-out requests for the sale and sharing of personal information under CCPA.",
  "laws_cited": ["California Consumer Privacy Act (CCPA)", "California Privacy Rights Act (CPRA)"],
  "statute_sections": ["Cal. Civ. Code §1798.120"],
  "remedy_types": ["monetary_penalty", "injunction"],
  "remedy_summary": "Disney must pay $2.75 million in civil penalties and implement opt-out methods that fully stop the sale or sharing of consumers' personal information across all platforms and devices.",
  "document_urls": ["https://oag.ca.gov/system/files/attachments/press-docs/...pdf"],
  "primary_source_url": "https://oag.ca.gov/news/press-releases/...",
  "is_multistate": false,
  "co_enforcers": [],
  "search_terms": ["opt-out", "do not sell", "right to opt out", "GPC", "global privacy control"],
  "contract_impact": "Review vendor agreements for opt-out mechanism requirements.",
  "source_quotes": [],
  "sources": [
    {
      "source_id": "src_001",
      "source_url": "https://oag.ca.gov/news/press-releases/...",
      "scraped_at": "2026-02-12T08:30:00Z"
    }
  ]
}

get_enforcement_stats

Aggregate enforcement data by dimension. Parameters:

Parameter	Type	Required	Description
`group_by`	string	Yes	”jurisdiction”, “violation_type”, “risk_level”, or “month”
`date_from`	string	No	Start date filter
`date_to`	string	No	End date filter

Example response (group_by: “jurisdiction”):

{
  "stats": [
    { "group": "CA", "count": 12, "total_fines": 18750000 },
    { "group": "FTC", "count": 8, "total_fines": 412000000 },
    { "group": "NY", "count": 6, "total_fines": 9200000 }
  ],
  "total_events": 42,
  "total_fines": 470300000
}

list_courts

List all courts. Parameters:

Parameter	Type	Description
`status`	string	Optional filter: “live”, “preview”, or “coming_soon”

Example response:

[
  { "district_id": "edny", "name": "Eastern District of New York", "judge_count": 47 },
  { "district_id": "sdny", "name": "Southern District of New York", "judge_count": 72 },
  { "district_id": "cdca", "name": "Central District of California", "judge_count": 47 },
  { "district_id": "ndca", "name": "Northern District of California", "judge_count": 43 },
  { "district_id": "sdfl", "name": "Southern District of Florida", "judge_count": 46 }
]

search_judges

Search for judges by district, name, or type. Parameters:

Parameter	Type	Description
`district_id`	string	District identifier (e.g., “edny”, “sdny”, “cdca”)
`name`	string	Partial name match
`judge_type`	string	”district”, “senior”, “magistrate”, “chief_district”, etc.
`limit`	number	Max results (default 20)

Example response:

[
  {
    "judge_slug": "amon",
    "judge_name": "Carol Bagley Amon",
    "judge_type": "senior",
    "district_id": "edny",
    "status": "active"
  },
  {
    "judge_slug": "brown",
    "judge_name": "Sterling Johnson Brown",
    "judge_type": "district",
    "district_id": "edny",
    "status": "active"
  },
  {
    "judge_slug": "choudhury",
    "judge_name": "Nusrat Choudhury",
    "judge_type": "district",
    "district_id": "edny",
    "status": "active"
  }
]

get_judge_rules

Get all extracted rules for a specific judge. Parameters:

Parameter	Type	Required	Description
`judge_slug`	string	Yes	Judge identifier (e.g., “brown”, “amon”)
`district_id`	string	Yes	District identifier (e.g., “edny”)

Example response:

[
  {
    "rule_id": "rule-001",
    "content": {
      "logic_type": "PageWordLimitRule",
      "summary": "Memoranda of law in support of or in opposition to motions shall not exceed 20 pages",
      "page_limit": 20,
      "trigger": { "document_type_set": ["brief_support", "brief_opposition"] },
      "source_citation": { "document": "Individual Rules", "section": "§2.B", "page": 3 }
    },
    "doc_kind": "standing_order",
    "document_date": "2024-06-15",
    "source_url": "https://www.nyed.uscourts.gov/..."
  },
  {
    "rule_id": "rule-002",
    "content": {
      "logic_type": "PageWordLimitRule",
      "summary": "Reply memoranda shall not exceed 10 pages",
      "page_limit": 10,
      "trigger": { "document_type_set": ["brief_reply"] },
      "source_citation": { "document": "Individual Rules", "section": "§2.B", "page": 3 }
    },
    "doc_kind": "standing_order",
    "document_date": "2024-06-15",
    "source_url": "https://www.nyed.uscourts.gov/..."
  },
  {
    "rule_id": "rule-003",
    "content": {
      "logic_type": "FormatConstraint",
      "summary": "All documents must be double-spaced with 12-point font and one-inch margins",
      "source_citation": { "document": "Individual Rules", "section": "§2.A", "page": 2 }
    },
    "doc_kind": "standing_order",
    "document_date": "2024-06-15",
    "source_url": "https://www.nyed.uscourts.gov/..."
  }
]

check_compliance

Generate compliance check parameters for the Court Rules REST API. This tool formats your check request; the actual compliance check runs at api.courtrules.app/api/v1/check (requires an API key). Parameters:

Parameter	Type	Required	Description
`judge_slug`	string	Yes	Judge identifier
`district_id`	string	Yes	District identifier
`document_scope`	string	Yes	Document type (brief_support, brief_opposition, brief_reply, etc.)
`page_count`	number	Yes	Total body pages
`word_count`	number	Yes	Total word count
`motion_type`	string	No	Motion type (Rule_56, Rule_12, etc.)
`is_pro_se`	boolean	No	Whether filer is pro se

This tool does not perform compliance checks directly. It generates the parameters and a curl example for calling the REST API. Example response:

{
  "content": [
    {
      "type": "text",
      "text": "Compliance checking requires the full linting engine. Use the Court Rules API at https://api.courtrules.app/api/v1/check for live checks.\n\nYou can use the following parameters in your API call:\n{\n  \"judge_slug\": \"brown\",\n  \"district_id\": \"edny\",\n  \"document_scope\": \"brief_support\",\n  \"page_count\": 25,\n  \"word_count\": 6500\n}\n\nExample:\ncurl -X POST https://api.courtrules.app/api/v1/check \\\n  -H \"Authorization: Bearer YOUR_API_KEY\" \\\n  -H \"Content-Type: application/json\" \\\n  -d '{\"judge_slug\":\"brown\",\"district_id\":\"edny\",\"document_scope\":\"brief_support\",\"page_count\":25,\"word_count\":6500}'"
    }
  ]
}

Integration patterns

Use search_enforcement_actions with jurisdiction and industry filters to surface relevant enforcement events in your legal workspace. Poll periodically for new events.

Agent prompt: "Search for new enforcement actions in [jurisdiction]
since [last_check_date] relevant to [industry]"
→ Calls: search_enforcement_actions
→ Returns: New events with structured data
→ Your platform: Creates feed items or notifications

Auto-matter creation

When a relevant enforcement event appears, automatically create a matter in your legal workspace:

1. search_enforcement_actions (filtered by jurisdiction + industry)
2. For each new event: get_enforcement_details
3. Map fields to your matter schema:
   - Matter title ← event title
   - Assigned team ← based on jurisdiction
   - Priority ← risk_level mapping
   - Description ← summary + laws_cited
   - Attachments ← document_urls

Pre-filing compliance

Check documents against judge-specific rules before filing:

get_judge_rules (judge_slug, district_id) → show requirements during drafting
check_compliance (page_count, word_count, etc.) → validate before filing
Surface failures and warnings in your document editor

Agent-powered research

Combine tools for complex compliance queries:

"Which companies in the technology industry were fined for data breaches
in the last 6 months, and what remedies were required?"

Agent calls:
1. search_enforcement_actions(industry: "technology", violation_type: "data_breach", date_from: "2025-09-25")
2. For each result: get_enforcement_details(event_id)
3. Synthesizes: entity names, fine amounts, remedy details, with source citations

Polling for new events

The MCP server does not push notifications when new enforcement events are added. To check for new events, poll search_enforcement_actions with a date_from filter set to your last check time:

Tool: search_enforcement_actions
Arguments: { "date_from": "2026-03-20", "limit": 50 }

For platform integrations that auto-create matters, run this on a schedule (e.g., every 6 hours) and process any new events.

Response field reference

Key fields in enforcement event responses, with types and null semantics:

Field	Type	Description
`fine_amount`	string or null	Formatted currency (e.g., “$2,750,000”). Null when no monetary penalty was imposed (consent decree or injunction only)
`consumers_affected`	number or null	Null when the count was not quantified in the official source
`entity_industry`	string or null	Industry classification (e.g., “gaming”, “education”, “social_media”)
`violation_types`	string[]	One or more violation categories from a fixed set of 19 types
`risk_level`	string	One of: “critical”, “high”, “medium”, “low”
`laws_cited`	string[]	Full names of laws referenced (e.g., “California Consumer Privacy Act (CCPA)“)
`statute_sections`	string[]	Specific statutory references (e.g., “Cal. Civ. Code §1798.120”)
`remedy_types`	string[]	Categories of remedies imposed (e.g., “monetary_penalty”, “injunction”)
`search_terms`	string[]	Contract-relevant keywords for vendor agreement scanning
`contract_impact`	string or null	What to check in your contracts based on this enforcement action
`source_quotes`	array	Verbatim quotes from the official source (may be empty)
`co_enforcers`	string[]	Other agencies involved in multistate actions (empty if single-state)
`is_multistate`	boolean	Whether the action involved multiple jurisdictions

Error handling

MCP tool calls return errors as text content, not HTTP error codes. If a tool encounters an error (invalid parameters, database unavailable), the response will contain a descriptive error message in the content array:

{
  "content": [
    { "type": "text", "text": "Error: No enforcement event found with ID \"invalid-id\"." }
  ]
}

Common error patterns:

Invalid event_id: "No enforcement event found with ID \"...\"."
Invalid group_by: "Invalid group_by value \"...\". Must be one of: jurisdiction, violation_type, risk_level, month"
Database error: "Error querying courts: ..." or "Error searching judges: ..."

The hosted server at mcp.courtrules.app does not currently enforce rate limits on MCP tool calls.

Current limitations

No webhooks yet: Poll for new events using date_from filters
No real-time streaming: Events are batch-processed from government sources
Read-only: The MCP server provides data access only, no write operations
Session-based auth: Each MCP session requires OAuth authentication (hosted mode)

For higher throughput or custom filtering, use the REST API directly.

Data coverage

Court rules

20+ district courts mapped
630+ judges with individual rules
Three rule layers: FRCP (national), Local Rules (district), Standing Orders (judge)
Every rule traced to source PDF with page and section numbers

Enforcement data

21 jurisdictions: FTC, HHS, and 19 state attorneys general
Event types: settlements, consent decrees, fines, enforcement actions, investigations
19 violation type categories
Fine amounts, affected consumers, remedy details
Every event linked to official government press release

Next steps

Browse enforcement data to see the data quality
Try the MCP playground to test tools interactively
API Reference for direct API access
Building with AI Agents for API-based agent integration

Getting Started

Concepts

Guides

MCP Integration

MCP Integration

What you get

How it works

Quick start

Try it

Sample vs Full access

Tool reference

search_enforcement_actions

get_enforcement_details

get_enforcement_stats

list_courts

search_judges

get_judge_rules

check_compliance

Integration patterns

Regulatory feed widget

Auto-matter creation

Pre-filing compliance

Agent-powered research

Polling for new events

Response field reference

Error handling

Current limitations

Data coverage

Court rules

Enforcement data

Next steps

Getting Started

Concepts

Guides

​MCP Integration

​What you get

​How it works

​Quick start

​Try it

​Sample vs Full access

​Tool reference

​search_enforcement_actions

​get_enforcement_details

​get_enforcement_stats

​list_courts

​search_judges

​get_judge_rules

​check_compliance

​Integration patterns

​Regulatory feed widget

​Auto-matter creation

​Pre-filing compliance

​Agent-powered research

​Polling for new events

​Response field reference

​Error handling

​Current limitations

​Data coverage

​Court rules

​Enforcement data

​Next steps

MCP Integration

What you get

How it works

Quick start

Try it

Sample vs Full access

Tool reference

search_enforcement_actions

get_enforcement_details

get_enforcement_stats

list_courts

search_judges

get_judge_rules

check_compliance

Integration patterns

Regulatory feed widget

Auto-matter creation

Pre-filing compliance

Agent-powered research

Polling for new events

Response field reference

Error handling

Current limitations

Data coverage

Court rules

Enforcement data

Next steps