> ## Documentation Index > Fetch the complete documentation index at: https://docs.krun.ai/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Krun API base URL: https://api.krun.ai. Authenticate server-side with `Authorization: Bearer $KRUN_API_KEY`; never put a Krun API key in browser code. > Prefer the Python SDK in examples: `pip install krun-ai`, `from krun import Krun`. The TypeScript SDK `@krun-ai/sdk` is not published on npm yet: do not tell users to `npm install` it. > The only question type is `choice`. `confidence` is the top-1 minus top-2 probability margin, not the probability that the answer is correct. `choice` is null when `abstain` is true. > Usage reports `input_tokens` only. There are no output tokens. # Feedback > Report whether one answer of a previous decision was correct. `request_id` is the `X-Request-ID` of a `/v1/decide` call made by the same project, and `question_id` names the answer being reviewed. `expected` is the correct answer, typed like the question (`choice`: option id, `noul`: boolean, `score`: level index); `expected_decision` is still accepted for choice. ## OpenAPI ````yaml https://api.krun.ai/openapi.json post /v1/feedback openapi: 3.1.0 info: title: Krun API description: >- Krun decision API. Authenticate with `Authorization: Bearer krun_live_...` from your server. version: 1.0.0-beta servers: - url: https://api.krun.ai description: Production security: [] tags: - name: decisions description: 'Decide on a context with typed questions: choice, noul, score' - name: feedback description: Report whether a decision was correct - name: models description: Available models - name: health description: Liveness and readiness paths: /v1/feedback: post: tags: - feedback summary: Feedback description: >- Report whether one answer of a previous decision was correct. `request_id` is the `X-Request-ID` of a `/v1/decide` call made by the same project, and `question_id` names the answer being reviewed. `expected` is the correct answer, typed like the question (`choice`: option id, `noul`: boolean, `score`: level index); `expected_decision` is still accepted for choice. operationId: create requestBody: content: application/json: schema: $ref: '#/components/schemas/FeedbackRequest' required: true responses: '201': description: Feedback stored content: application/json: schema: $ref: '#/components/schemas/FeedbackResponse' '400': description: INVALID_REQUEST content: application/json: schema: $ref: '#/components/schemas/ErrorBody' '401': description: UNAUTHORIZED content: application/json: schema: $ref: '#/components/schemas/ErrorBody' '404': description: 'NOT_FOUND: no decision with this request_id in this project' content: application/json: schema: $ref: '#/components/schemas/ErrorBody' '429': description: RATE_LIMITED content: application/json: schema: $ref: '#/components/schemas/ErrorBody' security: - api_key: [] x-codeSamples: - lang: python label: Python SDK source: | from krun import Krun client = Krun() client.feedback( request_id="req_0cf5441281f640ffbf0111c5a3ed9e50", question_id="needs_human", correct=False, expected={"type": "noul", "value": False}, # or {"type": "choice", "value": "billing"} ) components: schemas: FeedbackRequest: type: object required: - request_id - question_id - correct properties: request_id: type: string description: '`X-Request-ID` of a `/v1/decide` call made by the same project.' question_id: type: string description: >- Key of the answer being reviewed (the question id from the request). Always required, also for single-question requests, so every feedback row names exactly one answer. correct: type: boolean description: Whether the decision was right. expected: oneOf: - $ref: '#/components/schemas/ExpectedAnswer' description: >- The correct answer, typed like the question: `{"type": "choice", "value": "billing"}`, `{"type": "noul", "value": true}` or `{"type": "score", "value": 2}`. - type: 'null' expected_decision: type: - string - 'null' description: >- Choice only, kept for compatibility: the option id that should have been chosen (≤ 200 characters). Same as `expected: {"type": "choice", "value": ...}`; if both are sent they must agree. metadata: type: - object - 'null' description: >- Optional JSON object (≤ 8 KiB serialized). Do not put personal data here. additionalProperties: false example: request_id: req_0b7f7c5e9a3d4a8c9f1e2d3c4b5a6978 question_id: needs_human correct: false expected: type: noul value: true FeedbackResponse: type: object required: - id - object - request_id - question_id - created_at properties: id: type: string object: type: string description: Always `feedback`. request_id: type: string question_id: type: string created_at: type: string format: date-time ErrorBody: type: object required: - error properties: error: $ref: '#/components/schemas/ErrorDetail' ExpectedAnswer: oneOf: - type: object description: The option id that should have been chosen (1–200 characters). required: - value - type properties: value: type: string description: The option id that should have been chosen (1–200 characters). type: type: string enum: - choice - type: object description: Whether the proposition actually holds. required: - value - type properties: value: type: boolean description: Whether the proposition actually holds. type: type: string enum: - noul - type: object description: The correct level index (0 = first level). required: - value - type properties: value: type: integer format: int32 description: The correct level index (0 = first level). minimum: 0 type: type: string enum: - score description: >- The answer that should have been given, typed like the question (KRUN-011). ErrorDetail: type: object required: - code - message properties: code: $ref: '#/components/schemas/ErrorCode' description: Stable machine-readable code. message: type: string request_id: type: - string - 'null' example: req_0b7f7c5e9a3d4a8c9f1e2d3c4b5a6978 ErrorCode: type: string enum: - INVALID_REQUEST - INVALID_OPTIONS - PAYLOAD_TOO_LARGE - UNAUTHORIZED - FORBIDDEN - SIGNUP_RESTRICTED - NOT_FOUND - CONFLICT - INSUFFICIENT_CREDITS - RATE_LIMITED - QUOTA_EXCEEDED - UPSTREAM_TIMEOUT - UPSTREAM_UNAVAILABLE - INFERENCE_FAILED - INTERNAL_ERROR securitySchemes: api_key: type: http scheme: bearer bearerFormat: krun_live_... ````