Back

midpage caselaw
Description
Interact with US caselaw: Summarizes text, answers questions, and checks legal statements.
Details
https://midpage-plugin.onrender.com/.well-known/ai-plugin.json
{
"schema_version": "v1",
"name_for_model": "midpage_caselaw",
"name_for_human": "midpage caselaw",
"description_for_model": "**Overview of Capabilities (Allows users to interact with US case law)**:\n1. Generate and run queries for user searches to find cases, with optional date and jurisdiction filters.\n2. Given a case citation, access case excerpts, and provide quotes.\n3. Detect legal propositions, extract relevant quotes, and assess their veracity based on cited case excerpts.\n\n**Facts**:\n- Legal citations can be like 'ABC Corp. v. DEF Inc., 123 F.3d 456 (S.D.N.Y 2006)', 'ABC Corp., 123 F.3d at 569', '123 F.3d 456', or '123 F.3d 456, 569'.\n- A legal proposition is a statement of law with a case citation. An example is 'Mutual assent must be unambiguous. ABC Corp., 123 F.3d at 569'\n\n**Instructions:**\n- Steps for Detected Legal Propositions - ALWAYS IN ORDER: \n1. Access the Relevant Excerpts: Retrieve the relevant excerpts from the cited case(s) that pertain to the proposition in question.\n2. Analyze the Excerpts: Carefully review the provided excerpts to determine if they directly support or undermine the proposition.\n3. Provide a Confidence Rating:\n- High: Use this rating when there is direct support or contradiction in the excerpts for the proposition.\n- Low: Use this rating for all other situations, including when the support or contradiction is indirect, inferred, or not explicitly stated in the excerpts.\n4. Summarize the Assessment: Conclude by summarizing whether the proposition is supported or not supported based on the analysis and provide the confidence rating.\n- Steps for Provided Citations Without Legal Propositions - ALWAYS IN ORDER:\n1. NEVER summarize unless asked.\n2. Query text excerpts. If the user didn't specify what he wants to know, ask.\n3. Inform the user and ask specific questions about the case.\n\n**Query Instructions:**\n- Use search queries to find relevant cases on the topic of interest.\n- Once a case is identified, ask the user what specific information they want from that case.\n- Formulate a specific retrieval query based on the user's response to get the relevant excerpts from the case.\n\n**Output Instructions:**\n- Give intermediate answers between multiple requests.\n- To answer questions about a case, always get excerpts. Never base your conclusions on results of /search. \n- If confused about a case, seek clarification.\n- Provide accurate, verbatim quotations. Never cite unless it's in one of the case's text excerpts.\n- ALWAYS give the quote first, then assess the proposition.\n- Consider the 'additional_information' field in responses.\n- ALWAYS state that quotes are generated and should be user-reviewed.\n\n**About Midpage (the company behind this extension):**\nWe're a tech startup from New York and Berlin, making language models for lawyers. For midpage_caselaw usage, direct users to https://midpage.ai/how-to-use, then explain. Feedback goes to feedback@midpage.ai.",
"description_for_human": "Interact with US caselaw: Searches, summarizes, answers, and checks legal statements.",
"auth": {
"type": "oauth",
"client_url": "https://dev-xlbshnzwy6q5frgo.us.auth0.com/authorize",
"scope": "offline_access openid email",
"authorization_url": "https://dev-xlbshnzwy6q5frgo.us.auth0.com/oauth/token",
"authorization_content_type": "application/json",
"verification_tokens": {
"openai": "def18a599b50468dbd830204cfaf6545"
}
},
"api": {
"type": "openapi",
"url": "https://midpage-plugin.onrender.com/.well-known/openapi.yaml",
"has_user_authentication": false
},
"logo_url": "https://midpage-plugin.onrender.com/.well-known/midpage-icon.png",
"contact_email": "feedback@midpage.ai",
"legal_info_url": "https://www.midpage.ai/privacy-policy"
}
https://midpage-plugin.onrender.com/.well-known/openapi.yaml
openapi: 3.0.2
info:
title: midpage caselaw
description: Researching case law
version: 1.1.0
servers:
- url: https://midpage-plugin.onrender.com
paths:
/search:
post:
description: Returns a list of cases. Use this when the user wants to search for cases. Do not use this if you already know the citation(s) of requested cases.
operationId: search_post
requestBody:
content:
application/json:
schema:
$ref: "#/components/schemas/SearchRequest"
required: true
responses:
"200":
description: Successful Response
content:
application/json:
schema:
$ref: "#/components/schemas/SearchResult"
"422":
description: Validation Error
content:
application/json:
schema:
$ref: "#/components/schemas/HTTPValidationError"
/get_excerpts:
post:
description: Use this to get text excerpts for one or multiple documents. Use it to answer questions validate propositions or summarize cases. Use this over search if you know the case citations.
operationId: get_excerpts_post
requestBody:
content:
application/json:
schema:
$ref: "#/components/schemas/ExcerptRequest"
required: true
responses:
"200":
description: Successful Response
content:
application/json:
schema:
$ref: "#/components/schemas/DocumentExcerptList"
"404":
description: Not Found
content:
application/json:
schema:
$ref: "#/components/schemas/HTTPNotFound"
"422":
description: Validation Error
content:
application/json:
schema:
$ref: "#/components/schemas/HTTPValidationError"
components:
schemas:
ExcerptRequest:
title: ExcerptRequest
required:
- queryIdentifiers
type: object
description: A list of tuples, each tuple containing a retrievalQuery -- used to extract excerpts from the cases -- and the citation of a case.
properties:
queryIdentifiers:
title: Query and Identifiers
type: array
items:
type: object
required:
- retrievalQuery
- citation
properties:
retrievalQuery:
title: Query
type: string
description: A query that will be used to extract excerpts from a case. This could be a legal proposition or a question that should be answered.
citation:
title: Identifier
type: string
description: The citation of a case mentioned by the user, e.g., "123 F.3d 456". Don't use any document ids here.
SearchRequest:
title: SearchRequest
required:
- query
- jurisdictionType
type: object
description: Search for documents by a query that matches the user's request. It could match a proposition issue or text that appears verbatim in a case. Optionally, filter by date (format YYYY-MM-DD) and jurisdiction. If you don't have access to a filter, add it in the query.
properties:
query:
title: Query
type: string
description: The search query that matches the user's request. This could be a legal proposition or any text that appears verbatim in a case.
negativeQuery:
title: Negative Query
type: string
description: Optional secondary query that ranks cases lower if they match this query.
startDate:
title: Start Date
type: string
format: date
description: Start date (inclusive) for filtering cases. Format YYYY-MM-DD.
endDate:
title: End Date
type: string
format: date
description: End date (inclusive) for filtering cases. Format YYYY-MM-DD.
jurisdictionType:
title: Jurisdiction Type
type: string
enum: [ state, federal, state_and_federal ]
description: The type of courts to include in the search.
circuits:
title: Circuits
type: array
items:
type: string
enum: [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, dc, federal, supreme_court ]
description: The circuits to include. This is only applied to federal cases.
states:
title: States
type: array
items:
type: string
description: The name of the states, e.g. "New York", "Ohio" to include -- no abbreviations.
includeCourts:
title: List of courts to include.
type: array
items:
type: string
description: An additional list of courts to narrow down the search to. This is applied on top of other filters.
Document:
title: Document
required:
- text
type: object
description: The full text of specific case detected, and the user's legal proposition (if any).
properties:
document_id:
title: Id
type: string
text:
title: Text
type: string
metadata:
$ref: "#/components/schemas/DocumentMetadata"
proposition:
title: Proposition
type: string
SearchResult:
title: SearchResult
required:
- results
type: object
description: A list of cases responsive to the search request.
properties:
results:
title: Search results
type: array
items:
$ref: "#/components/schemas/Document"
DocumentExcerptList:
title: DocumentExcerptList
required:
- excerpts
type: object
description: A list of DocumentExcerpts.
properties:
excerpts:
title: DocumentExcerpts
type: array
items:
$ref: "#/components/schemas/DocumentExcerpts"
DocumentExcerpts:
title: DocumentExcerpts
required:
- excerpts
type: object
description: Metadata and excerpts from a case.
properties:
metadata:
$ref: "#/components/schemas/DocumentMetadata"
excerpts:
title: Excerpts
type: array
items:
$ref: "#/components/schemas/Excerpt"
Excerpt:
title: Excerpt
required:
- text
type: object
description: An excerpt from a case and a link to the quote in the document.
properties:
text:
title: Text
type: string
href:
title: href
type: string
description: A URL pointing to the case and the specific quote. Show this to the user.
DocumentMetadata:
title: DocumentMetadata
type: object
description: The metadata associated with a document.
properties:
url:
title: Url
type: string
title:
title: The title / case name
type: string
court:
title: Court
type: string
filed_date:
title: Filed Date
type: string
reporter:
title: Reporter
type: string
cited_as:
title: How the case is cited
type: string
jurisdiction:
title: Jurisdiction
type: string
judge_name:
title: Judge
type: string
additional_information:
title: Additional Information
type: string
description: Additional information about the case, e.g., how often it has been cited or if it was overruled.
HTTPValidationError:
title: HTTPValidationError
type: object
properties:
detail:
title: Detail
type: array
items:
$ref: "#/components/schemas/ValidationError"
HTTPNotFound:
title: HTTPNotFound
type: object
properties:
detail:
title: Detail
type: string
description: Description of the resource not found.
ValidationError:
title: ValidationError
required:
- loc
- msg
- type
type: object
properties:
loc:
title: Location
type: array
items:
anyOf:
- type: string
- type: integer
msg:
title: Message
type: string
type:
title: Error Type
type: string
Discover other plugins from the legal category

Docket Alarm Plugin
Litigation research and analytics platform. Search, count, and track litigation across the Federal and State Courts
0 Comments

LawyerPR
Matching your ideal lawyer, in Japan. Let's Start with a Preliminary Review.
0 Comments

California Law
Get up to date access to California law.
0 Comments

US Federal Law
Get up to date access to Federal law.
0 Comments

Talk Law Brazil
Interact with laws and bills in progress in the Brazilian Congress.
0 Comments

LegalQA
Answer any inquiries related to Chinese law.
0 Comments

Halifax Bylaws
Your Halifax Bylaw's Expert.
0 Comments