SageOx LLM API

openapi: 3.1.0
info:
  title: SageOx Admin LLM API
  version: 1.0.0
  description: "# API Reference\n\n## Overview\n\nSageOx is a platform that captures team knowledge from discussions, decisions, and work context into a Ledger (per-repo historical record) and Team Context (team-wide shared knowledge). The SageOx API provides programmatic access to manage repositories, recordings, notifications, and team data with high performance and reliability.\n\n## Base URLs\n\n| Environment | URL |\n|---|---|\n| Local Development | `http://localhost:3000` |\n| Test | `https://test.sageox.ai` |\n| Production | `https://sageox.ai` |\n\n## Authentication\n\nAll requests to the SageOx API require authentication via JWT tokens issued by the Better Auth service.\n\nInclude your token in the `Authorization` header:\n```\nAuthorization: Bearer <token>\n```\n\n**Initial Auth Flow (CLI)**\n- CLI initiates device flow to obtain initial credentials\n- Exchange device code for JWT token\n- Store token securely for subsequent API requests\n- Refresh token when expired\n\n**Authenticated Endpoints**\n- Pass JWT in `Authorization: Bearer <token>` header\n- Tokens contain user identity and team membership\n- Invalid or expired tokens return 401 Unauthorized\n\n## Response Format\n\n### Success\nStandard response format with optional pagination metadata:\n```json\n{\n  \"success\": true,\n  \"data\": { ... }\n}\n```\n\n### Error\nAll errors follow a consistent format:\n```json\n{\n  \"success\": false,\n  \"error\": \"Human-readable error description\"\n}\n```\n\n**Status Codes**\n- `400` — Bad Request: Invalid parameters or malformed request\n- `401` — Unauthorized: Missing or invalid authentication token\n- `403` — Forbidden: Insufficient permissions for this resource\n- `404` — Not Found: Resource does not exist\n- `409` — Conflict: Request conflicts with current state (e.g., duplicate)\n- `429` — Rate Limited: Too many requests; retry with backoff\n- `500` — Internal Error: Server error; contact support if persistent\n\n## Pagination\n\nList endpoints support cursor-based pagination via query parameters:\n\n**Query Parameters**\n- `limit` — Number of results per page (default: 20, max: 100)\n- `offset` — Number of results to skip (default: 0)\n\n**Response Metadata**\nList responses include pagination info:\n```json\n{\n  \"success\": true,\n  \"data\": [ ... ],\n  \"meta\": {\n    \"total\": 150,\n    \"limit\": 20,\n    \"offset\": 0,\n    \"hasMore\": true\n  }\n}\n```\n\n## ID Formats\n\nResources use standardized ID formats for easy identification:\n\n| Resource | Format | Length | Example |\n|---|---|---|---|\n| Team | `team_<cuid2>` | 10 chars | `team_abc1234567` |\n| User | `usr_<cuid2>` | 10 chars | `usr_xyz9876543` |\n| Repository | `repo_<uuidv7>` | 36 chars | `repo_01934f5a-8b9c-7def-abcd-0123456789ab` |\n| Recording | `rec_<uuidv7>` | 36 chars | `rec_01934f5a-8b9c-7def-abcd-0123456789ab` |\n| Image | `img_<uuidv7>` | 36 chars | `img_01934f5a-8b9c-7def-abcd-0123456789ab` |\n| Notification | `notif_<cuid2>` | 14 chars | `notif_ab1cd2ef3g` |\n\nUse these IDs for all API operations. IDs are unique across environments.\n\n## Rate Limiting\n\nThe API applies rate limiting to protect against abuse and ensure fair access:\n\n**Authenticated Endpoints**\n- Limited per user: depends on subscription tier\n- Quota resets hourly\n\n**Unauthenticated Endpoints**\n- Limited per IP address\n- More restrictive than authenticated limits\n\nWhen rate limited, the API returns `429 Too Many Requests`. Retry requests with exponential backoff:\n```\nwait = min(2^attempt * 100ms, 10s)\n```\n\n## Timestamps\n\nAll timestamps in API responses use RFC 3339 / ISO 8601 format in UTC:\n```\n2025-12-03T10:30:00Z\n```\n\nUse this format when providing timestamps in requests as well. The API rejects timestamps in other formats.\n\n## SDKs & Tools\n\n- **OpenAPI Spec** — Full schema available at `/api/openapi.json`\n- **Postman Collection** — Import from `/api/postman.json` for interactive exploration\n- **CLI** — Use `ox` CLI for command-line access\n"
  contact:
    name: SageOx Team
  license:
    name: MIT
servers:
- url: http://localhost:3000
  description: Devcontainer
- url: https://test.sageox.ai
  description: Test
- url: https://sageox.ai
  description: Production
security:
- bearerAuth: []
tags:
- name: LLM
  description: 'OpenAI-compatible chat completion endpoints. Proxies requests to configured LLM providers (Bedrock, OpenAI) with model routing, token usage tracking, and streaming support. Follows the OpenAI `/v1/chat/completions` format.

    '
paths:
  /v1/models:
    get:
      operationId: listModels
      summary: List available models
      description: List all available LLM models that can be used for chat completions
      tags:
      - LLM
      security:
      - bearerAuth: []
      responses:
        '200':
          description: List of available models
          content:
            application/json:
              schema:
                type: object
                required:
                - object
                - data
                properties:
                  object:
                    type: string
                    enum:
                    - list
                    example: list
                  data:
                    type: array
                    items:
                      type: object
                      required:
                      - id
                      - object
                      - created
                      - owned_by
                      properties:
                        id:
                          type: string
                          description: Model identifier
                          example: gpt-4
                        object:
                          type: string
                          enum:
                          - model
                          example: model
                        created:
                          type: integer
                          description: Unix timestamp of when the model was created
                          example: 1677610602
                        owned_by:
                          type: string
                          description: Organization that owns the model
                          example: sageox
        '401':
          description: Unauthorized - Invalid or missing bearer token
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
        '500':
          description: Internal server error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
  /v1/chat/completions:
    post:
      operationId: createChatCompletion
      summary: Create chat completion
      description: 'Create a chat completion using OpenAI-compatible API format.

        Supports both streaming and non-streaming responses.


        For streaming responses, set `stream: true` in the request body.

        The response will be a text/event-stream with Server-Sent Events (SSE).

        '
      tags:
      - LLM
      security:
      - bearerAuth: []
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              required:
              - model
              - messages
              properties:
                model:
                  type: string
                  description: ID of the model to use
                  example: gpt-4
                messages:
                  type: array
                  description: List of messages in the conversation
                  minItems: 1
                  items:
                    type: object
                    required:
                    - role
                    - content
                    properties:
                      role:
                        type: string
                        enum:
                        - system
                        - user
                        - assistant
                        description: Role of the message author
                        example: user
                      content:
                        type: string
                        description: Content of the message
                        example: Hello, how are you?
                stream:
                  type: boolean
                  description: Whether to stream the response using SSE
                  default: false
                  example: false
                temperature:
                  type: number
                  description: Sampling temperature between 0 and 2
                  minimum: 0
                  maximum: 2
                  default: 1
                  example: 0.7
                max_tokens:
                  type: integer
                  description: Maximum number of tokens to generate
                  minimum: 1
                  example: 1000
                top_p:
                  type: number
                  description: Nucleus sampling parameter
                  minimum: 0
                  maximum: 1
                  example: 1
                frequency_penalty:
                  type: number
                  description: Frequency penalty between -2.0 and 2.0
                  minimum: -2
                  maximum: 2
                  default: 0
                  example: 0
                presence_penalty:
                  type: number
                  description: Presence penalty between -2.0 and 2.0
                  minimum: -2
                  maximum: 2
                  default: 0
                  example: 0
                stop:
                  oneOf:
                  - type: string
                  - type: array
                    items:
                      type: string
                  description: Up to 4 sequences where the API will stop generating
                  example:
                  - \n
                n:
                  type: integer
                  description: Number of completions to generate
                  minimum: 1
                  maximum: 10
                  default: 1
                  example: 1
                user:
                  type: string
                  description: Unique identifier for the end-user
                  example: user-12345
      responses:
        '200':
          description: Chat completion response
          content:
            application/json:
              schema:
                type: object
                required:
                - id
                - object
                - created
                - model
                - choices
                - usage
                properties:
                  id:
                    type: string
                    description: Unique identifier for the completion
                    example: chatcmpl-abc123
                  object:
                    type: string
                    enum:
                    - chat.completion
                    example: chat.completion
                  created:
                    type: integer
                    description: Unix timestamp of when the completion was created
                    example: 1677858242
                  model:
                    type: string
                    description: Model used for the completion
                    example: gpt-4
                  choices:
                    type: array
                    description: List of completion choices
                    items:
                      type: object
                      required:
                      - index
                      - message
                      - finish_reason
                      properties:
                        index:
                          type: integer
                          description: Index of the choice
                          example: 0
                        message:
                          type: object
                          required:
                          - role
                          - content
                          properties:
                            role:
                              type: string
                              enum:
                              - assistant
                              example: assistant
                            content:
                              type: string
                              description: Generated message content
                              example: Hello! I'm doing well, thank you for asking.
                        finish_reason:
                          type: string
                          enum:
                          - stop
                          - length
                          - content_filter
                          - null
                          description: Reason why the completion finished
                          example: stop
                  usage:
                    type: object
                    required:
                    - prompt_tokens
                    - completion_tokens
                    - total_tokens
                    properties:
                      prompt_tokens:
                        type: integer
                        description: Number of tokens in the prompt
                        example: 10
                      completion_tokens:
                        type: integer
                        description: Number of tokens in the completion
                        example: 12
                      total_tokens:
                        type: integer
                        description: Total number of tokens used
                        example: 22
            text/event-stream:
              schema:
                type: string
                description: 'Server-Sent Events stream (when stream=true).

                  Each event is a JSON object representing a chunk of the completion.


                  Format:

                  data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-4","choices":[{"index":0,"delta":{"role":"assistant","content":"Hello"},"finish_reason":null}]}


                  The stream ends with:

                  data: [DONE]

                  '
              example: 'data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-4","choices":[{"index":0,"delta":{"role":"assistant","content":""},"finish_reason":null}]}


                data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-4","choices":[{"index":0,"delta":{"content":"Hello"},"finish_reason":null}]}


                data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-4","choices":[{"index":0,"delta":{"content":"!"},"finish_reason":null}]}


                data: {"id":"chatcmpl-abc123","object":"chat.completion.chunk","created":1677858242,"model":"gpt-4","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}


                data: [DONE]

                '
        '400':
          description: Bad request - Invalid parameters
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
              example:
                error:
                  message: Invalid request body
                  type: invalid_request_error
                  code: invalid_request
        '401':
          description: Unauthorized - Invalid or missing bearer token
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
              example:
                error:
                  message: Invalid authentication
                  type: invalid_request_error
                  code: invalid_api_key
        '429':
          description: Rate limit exceeded
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
              example:
                error:
                  message: Rate limit exceeded
                  type: rate_limit_error
                  code: rate_limit_exceeded
        '500':
          description: Internal server error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OpenAIError'
              example:
                error:
                  message: Internal server error
                  type: server_error
                  code: server_error
components:
  schemas:
    OpenAIError:
      type: object
      description: Error response following the OpenAI error format. Returned by LLM endpoints.
      required:
      - error
      properties:
        error:
          type: object
          description: Error details object.
          required:
          - message
          - type
          properties:
            message:
              type: string
              description: Human-readable error description.
              example: Invalid API key
            type:
              type: string
              description: 'Error category. Common values: `invalid_request_error`, `authentication_error`, `rate_limit_error`.'
              example: invalid_request_error
            code:
              type: string
              nullable: true
              description: Machine-readable error code for programmatic handling. `null` when no specific code applies.
              example: invalid_api_key
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: JWT
      description: 'JWT token obtained from the auth service (/api/auth/token).

        Token is validated using JWKS from the auth service.

        Required claims: sub (user_id), email, name, tier.

        '