Connect SchoolsReg to your AI workspace

Before You Start

• 1.Request access and wait for approval.
• 2.Open the approval email and copy your personal API token.
• 3.Keep the token private. It becomes the ChatGPT Action bearer key.

Step-by-Step Setup

1. 1.In ChatGPT, open Explore GPTs, then choose Create.
2. 2.Open the Configure tab. Fill in the GPT name and description.
3. 3.In Instructions, paste the instruction text from this page.
4. 4.Under Actions, choose Create new action.
5. 5.Set Authentication to API Key. If ChatGPT asks for the auth type, choose Bearer.
6. 6.Paste your approved SchoolsReg token into the API key field. Do not paste it into the YAML schema.
7. 7.Paste the OpenAPI schema below into the schema editor and save the action.
8. 8.Use the built-in test prompt, then ask a question such as: What does the agreement say about NEAB dues?

You are a SchoolsReg legal document assistant.

When the user asks about school legal documents, contracts, policies, regulations, dues, payroll deductions, grievances, benefits, transfers, leave, salaries, or agreement terms, call searchLegalDocuments.

Use query_text for the user's full question. Include as_of only when the user explicitly specifies a date. Summarize the returned answer clearly and cite the returned document titles. If no relevant result is returned, say that the SchoolsReg database did not return a matching excerpt.

openapi: 3.1.0
info:
  title: SchoolsReg Legal Search API
  description: Search ingested school legal and policy documents using semantic search. Use this action when the user asks about school agreements, regulations, policies, dues, benefits, transfers, leave, salaries, or other legal document content.
  version: 1.0.0
servers:
  - url: https://schoolsreg.com
    description: Production server
paths:
  /api/action/search:
    post:
      operationId: searchLegalDocuments
      summary: Search school legal documents
      description: Searches the SchoolsReg vector database for relevant school legal or policy document chunks and returns source titles plus excerpts.
      security:
        - bearerAuth: []
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: "#/components/schemas/SearchRequest"
      responses:
        "200":
          description: Relevant legal document chunks were returned.
          content:
            application/json:
              schema:
                $ref: "#/components/schemas/SearchResponse"
        "401":
          description: Missing API key.
        "403":
          description: Invalid API key.
        "422":
          description: Invalid request parameters.
components:
  schemas:
    SearchRequest:
      type: object
      additionalProperties: false
      properties:
        query_text:
          type: string
          description: The user question to search for in the legal document corpus.
        as_of:
          type: string
          format: date-time
          description: Optional ISO-8601 date/time for time-scoped search. If omitted, the current date/time is used.
      required:
        - query_text
    SearchResponse:
      type: object
      properties:
        as_of:
          type: string
          format: date-time
          description: The timestamp used for the time-scoped search.
        answer:
          type: string
          description: A readable summary of matched source chunks, including source titles.
        results:
          type: array
          items:
            $ref: "#/components/schemas/SearchResult"
    SearchResult:
      type: object
      properties:
        title:
          type: string
          description: Source document filename or title.
        content:
          type: string
          description: Matching source text excerpt.
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer