> ## Documentation Index > Fetch the complete documentation index at: https://docs.qode.world/llms.txt > Use this file to discover all available pages before exploring further. # Get interview details > Retrieves detailed results, transcripts, and assets for a specific interview session. ## OpenAPI ````yaml /api-reference/specs/integration.yaml get /integrations/candidates/interviews/{interview_id} openapi: 3.0.3 info: title: Integration Document version: 1.0.0 description: API endpoints to ingest candidate data to Qode. x-mint: mcp: enabled: true servers: - url: https://turing.gomu.ai/api/v1/ description: Staging - url: https://turing.qode.world/api/v1/ description: Production security: - bearerAuth: [] paths: /integrations/candidates/interviews/{interview_id}: get: tags: - Interviews summary: Get interview details description: >- Retrieves detailed results, transcripts, and assets for a specific interview session. operationId: getInterview parameters: - name: interview_id in: path required: true schema: type: string example: >- 35bdbf86-e40a-4c2b-ddaf-012318cfc630_5d1914ff-1574-4aa6-a017-60f927f7740a - name: session_id in: query required: false schema: type: string example: 5d1914ff-1574-4aa6-a017-60f927f7740a responses: '200': description: Interview details retrieved successfully headers: X-TraceID: description: Trace identifier for debugging and audit tracking schema: type: string example: 3915368fcce2042c3ff1eb947a6a5312 content: application/json: schema: $ref: '#/components/schemas/GetInterViewResponse' security: - bearerAuth: [] components: schemas: GetInterViewResponse: type: object properties: role_title: type: string example: Software Engineer mapped_role_title: type: string description: >- Mapped role title derived on the role title and interview transcript. example: Backend Developer candidate_email: type: string example: example@gmail.com candidate_id: type: string example: 5d19f4ff-1574-4aa6-a017-60f927f7740a interview_id: type: string example: >- 123cb1b6-0cc1-4ab6-aa3c-5e0c75da6c6f_5d19f4ff-1574-4aa6-a017-60f927f7740a iframe_url: type: string format: uri example: https://turing.qode.world/interview/fbSGW request_metadata: type: object nullable: true description: Custom metadata object passed during interview creation. interview_type: type: string nullable: true description: Type of AI interviewer used for this interview. enum: - TURING_STEM - TURING_NON_STEM - DOCKER_INTERVIEW - SOFTWARE_ENGINEER example: TURING_STEM sub_interview_type: type: string nullable: true description: >- Grouped job title for TURING_STEM and TURING_NON_STEM interview types. example: Project Manager session_id: type: string nullable: true description: ID of an interview session example: 4e9e55ee-63f3-4c04-87ec-82a1823a538d summary: type: object nullable: true properties: skill_summaries: type: array items: type: object properties: skill_name: type: string example: Quantitative Analysis and Statistical Techniques summary: type: string example: >- Demonstrated strong proficiency in quantitative analysis, particularly with financial data... highlights: type: array items: type: object properties: highlight_name: type: string example: Nuanced Model Comparison description: type: string example: >- Provided a clear and insightful comparison between statistical and ML models... areas_of_expertise: type: array items: type: object properties: skill_name: type: string example: Credit Risk Modeling interview_transcript_line: type: array items: type: object properties: speaker_name: type: string example: Turing transcript: type: string example: Hello, my name is Tracy... assets: type: object nullable: true properties: pdf_report_url: type: string format: uri example: >- https://cdn.qode.gg/assessment-pdf-turing/123cb1b6-0cc1-4ab6-aa3c-5e0c75da6c6f_assessment_profile.pdf video_url: type: string format: uri example: >- https://cdn.qode.gg/recording/tpb6JAKot5e7ldzZgu19/video_record.mp4 transcript_url: type: string format: uri example: https://cdn.qode.gg/recording/tpb6JAKot5e7ldzZgu19/chapter.json metadata: type: object nullable: true properties: drop_off_time: type: string format: date-time example: '2025-11-07T13:51:42.740Z' scheduled_time: type: string format: date-time example: '2025-11-07T13:51:42.740Z' start_time: type: string format: date-time example: '2025-11-07T13:51:42.740Z' end_time: type: string format: date-time example: '2025-11-07T15:10:43.000Z' scorecard_time: type: string format: date-time nullable: true description: Timestamp when the report is generated. example: '2025-11-07T15:12:00.000Z' realtime_status: type: string description: The realtime interview status, derived from the event stream. example: INTERVIEW_STARTED enum: - INTERVIEW_SCHEDULED - INTERVIEW_CREATED - INTERVIEW_GETTING_READY - INTERVIEW_STARTED - INTERVIEW_LOCKED - INTERVIEW_COMPLETED - INTERVIEW_TECH_FAILURE - INTERVIEW_ABANDONED - INTERVIEW_TIMEOUT - INTERVIEW_CONNECTION_FAILURE is_terminal: type: boolean description: Boolean indicating whether the interview can progress further. example: false duration: type: number nullable: true description: Duration of the interview in seconds example: 1800 interview_signals: $ref: '#/components/schemas/InterviewSignals' black_screen_timestamps: type: array description: Time ranges where a black screen was detected during the interview. items: type: object properties: start: type: string description: Start timestamp of the black screen period (HH:MM:SS.mmm). example: '00:00:01.600' end: type: string description: End timestamp of the black screen period (HH:MM:SS.mmm). example: '00:00:03.000' duration: type: number description: Duration of the black screen period in seconds. example: 1.4 default: [] example: - start: '00:00:01.600' end: '00:00:03.000' duration: 1.4 - start: '00:01:42.711' end: '00:01:48.411' duration: 5.7 - start: '00:01:52.878' end: '00:01:53.911' duration: 1.033 - start: '00:09:42.366' end: '00:09:48.533' duration: 6.167 - start: '00:09:49.343' end: '00:09:51.043' duration: 1.7 meeting_data: type: object properties: completed_status: type: string description: Completion status of the candidate's interview example: COMPLETED enum: - COMPLETED - PARTIALLY_COMPLETE - INCOMPLETED start_date_time: type: string format: date-time example: '2025-11-07T13:51:42.740Z' end_date_time: type: string format: date-time example: '2025-11-07T15:10:43.000Z' chapters: type: array description: List of chapters in the interview items: type: object properties: title: type: string nullable: true description: Title of the chapter start_time: type: string nullable: true format: date-time description: Start time of the chapter example: '2025-11-07T13:51:42.740Z' end_time: type: string nullable: true format: date-time description: End time of the chapter example: '2025-11-07T15:10:43.000Z' question_summary: type: array description: >- Summary of the questions answered by the candidate and its scoring by question and skill level. items: type: object properties: question: type: string description: Question text summary: type: string description: Summary of the candidate's answer timestamp: type: string nullable: true description: Timestamp of the question in the video recording question_score: type: object description: Scoring information of this question properties: agent_summary: type: string description: The agent's summary of the agent scoring information score: type: object nullable: true properties: score: type: number description: Score of the question ci_lower: type: number nullable: true description: >- Lower bound of the confidence interval in question level ci_upper: type: number nullable: true description: >- Upper bound of the confidence interval in question level skills_score: type: array nullable: true description: Scoring information of different skills in the question items: type: object properties: name: type: string description: Name of the skill score: type: object properties: score: type: number description: Score of the question ci_lower: type: number nullable: true description: >- Lower bound of the confidence interval in question level ci_upper: type: number nullable: true description: >- Upper bound of the confidence interval in question level summary: type: string description: Summary of the skill scoring information agents: type: array items: type: object properties: name: type: string description: Name of the agent summary: type: string description: Summary of the agent scoring information raw_transcript: type: string nullable: true description: Full raw transcript of the interview session. candidate_profile: type: object properties: avatar: type: string nullable: true format: uri description: URL of the candidate's avatar example: >- https://cdn.qode.gg/avatar/123cb1b6-0cc1-4ab6-aa3c-5e0c75da6c6f.png fullname: type: string nullable: true description: Name of the candidate example: Alan Turing year_of_experience: type: number nullable: true description: Number of years of experience of the candidate agent_response_summary: type: array nullable: true description: > Multi-agent assessment results for each interview question. See [Multi-Agent Assessment](/api-reference/multi-agent-assessment) for detailed schema documentation. items: $ref: '#/components/schemas/AgentResponseSummaryItem' overall_score: type: number nullable: true description: Aggregated score across all questions and agents (0–100 scale). example: 70 interview_result: type: string description: > Overall interview result (e.g., "Pass", "Fail", "Not enough interviews (minimum 10 required) for this role to evaluate Pass/Fail."). example: Fail count_all_core_domains: type: integer nullable: true description: Total number of core domains assessed in this interview. example: 2 count_covered_core_domains: type: integer nullable: true description: Number of core domains that were covered during the interview. example: 2 uncovered_domains: type: array description: >- List of domains that were not covered during the interview for that role title. example: - Data & DB - DevOps language_proficiency_scores: type: object description: >- Language proficiency assessment scores (Only available for BUSINESS_ANALYST, BUSINESS_ANALYST_MCQ and MULTI_BUSINESS_ANALYST) properties: coherence_score: type: number description: Coherence score for the assessed language. fluency_score: type: number description: Fluency score for the assessed language. grammar_score: type: number description: Grammar score for the assessed language. pronunciation_score: type: number description: Pronunciation score for the assessed language. prosody_score: type: number description: Prosody score for the assessed language. accuracy_score: type: number description: Accuracy score for the assessed language. completeness_score: type: number description: Completeness score for the assessed language. summary: type: string description: Summary of the language proficiency assessment. vocabulary_score: type: number description: Vocabulary score for the assessed language. language: type: string description: ISO or display code for the assessed language. other_languages: type: array description: Proficiency scores for additional languages. items: $ref: '#/components/schemas/LanguageProficiencyScoreItem' required: - coherence_score - fluency_score - grammar_score - pronunciation_score - summary - vocabulary_score code_task_result: type: object nullable: true description: >- Result of code task assessment including individual tasks and overall coding score. (Only for SOFTWARE_ENGINEER_CODING) properties: tasks: type: array description: List of code tasks with scores and pass counts. items: type: object properties: name: type: string description: Name of the code task. language_name: type: string description: Programming language used for the task. testcase_passrate: type: string nullable: true description: >- Ratio of passed test cases to total test cases for the submission. Value ranges from 0 to 1. code_execution_time: type: number nullable: true description: >- Average execution time in seconds across all test cases returned by the code execution engine. implementation_time: type: number nullable: true description: >- Time in seconds taken by the user to implement the solution, measured from task start to code submission. ram_consumption: type: number nullable: true description: >- Average memory usage in kilobytes (KB) across all executed test cases. algorithm_complexity: type: string nullable: true description: >- Estimated algorithmic time complexity of the submitted code (e.g., O(1), O(n), O(n log n), O(n²)), determined through static code analysis. status: type: string description: Status of the task (Passed or Failed). enum: - Passed - Failed required: - name - score - passed_count - total_count coding_score: type: number description: Overall coding score across all tasks. required: - tasks - coding_score card_check_image_path: type: string nullable: true description: Path to the identity screenshot demonstrated_skills_across_domains: type: array nullable: true description: >- Skills demonstrated across different domains with confidence intervals and stability metrics. items: type: object properties: domain_name: type: string example: ML Modeling & Algorithms domain_category: type: string example: Core observations: type: integer description: Number of observations for this domain. example: 7 domain_weight: type: number description: Weight of this domain in the overall assessment. example: 0.23 ci_lower: type: integer description: Lower bound of the confidence interval. example: 68 ci_upper: type: integer description: Upper bound of the confidence interval. example: 69 stability: type: string description: Stability label for the confidence interval. example: Highly Reliable mean_score_across_skills: type: integer description: Mean score across all skills in this domain. example: 69 skills: type: array description: List of skill names in this domain. items: type: string example: - Feature Engineering - Model Selection - Model Training scored_skills: type: array nullable: true description: >- Individual skills with their pooled mean scores and confidence intervals. items: type: object properties: skill_name: type: string example: Zapier pooled_mean_score: type: integer description: Pooled mean score for this skill (0-100 scale). example: 67 ci_lower: type: integer description: Lower bound of the confidence interval. example: 66 ci_upper: type: integer description: Upper bound of the confidence interval. example: 68 stability: type: string description: Stability label for the confidence interval. example: Highly Reliable interview_level_bootstrap: type: object nullable: true description: >- Bootstrap confidence interval and interpretation for the overall interview score. properties: ci_lower: type: integer description: >- Lower bound of the confidence interval for the overall interview score. example: 51 ci_upper: type: integer description: >- Upper bound of the confidence interval for the overall interview score. example: 63 label: type: string description: Stability label based on the confidence interval width. example: Moderately Uncertain interpretation: type: string description: Human-readable interpretation of the stability. example: >- Some questions were scored differently, making the interview outcome less precise but still meaningful. grading_rubric_result: type: object nullable: true description: Result of the grading rubric assessment. properties: criteria: type: array description: List of rubric criteria with name, score, and weight. items: type: object properties: name: type: string example: Resume & Technical deep-dive interview with Tracy score: type: number example: 0 weight: type: number example: 0.6 assessment_version: type: string description: Version of the assessment. nullable: true enum: - no_ci_v1 - ci_v2 - pf_domain_v3 - pf_domain_3p_v4 latency_summary: type: object nullable: true description: Latency analysis summary for the interview session. properties: turn_details: type: array description: Per-turn latency breakdown. items: type: object properties: turn: type: number description: Turn number in the conversation. example: 1 speaker: type: string description: Speaker identifier for this turn (Tracy, Candidate). example: Tracy latency: type: number nullable: true description: >- Response latency in seconds for this turn. (Tracy turn latency will also be null) example: 2.5 content: type: string description: Content of the turn. example: Hello, my name is Tracy. start_time: type: string description: Start time of the turn. example: '00:01:49.947' end_time: type: string description: End time of the turn. example: '00:01:51.567' median: type: number description: Median latency across all turns in seconds. example: 2.3 avg_raw: type: number description: Raw average latency across all turns in seconds. example: 3.1 avg_clean: type: number description: Average latency excluding anomaly turns in seconds. example: 2.4 anomaly_count: type: number description: >- Number of turns flagged as latency anomalies. (Anomaly turn is 30 seconds or more) example: 0 anomaly_turns: type: array description: Turn numbers flagged as latency anomalies. items: type: number example: [] event: type: string example: started enum: - SCHEDULED - STARTED - NO_SHOW - CANDIDATE_DROP_OFF - TECHNICAL_GLITCH - INTERVIEW_COMPLETED - REPORT_GENERATED timestamp: type: string example: '2025-11-11T08:04:04Z' description: event timestamp in ISO 8601 format example: role_title: Software Engineer mapped_role_title: Backend Developer candidate_email: example@gmail.com candidate_id: f60c8c7b-a3a6-4d2f-af0c-8950cc437ddd interview_id: >- c912a118-08d8-49aa-bf9f-602ec5aea395_f60c8c7b-a3a6-4d2f-af0c-8950cc437ddd iframe_url: https://turing.gomu.ai/interview/fbSGW interview_type: TURING_STEM sub_interview_type: Software Engineer request_metadata: {} session_id: 4e9e55ee-63f3-4c04-87ec-82a1823a538d candidate_profile: avatar: null fullname: '' year_of_experience: null summary: skill_summaries: - skill_name: Technical Knowledge summary: >- The candidate mentioned experience with front-end, back-end, and database design. highlights: - highlight_name: Experience with Data Annotation description: >- The candidate has some experience with data annotation for a vehicle dataset. areas_of_expertise: - skill_name: Full-stack development - skill_name: Kafka - skill_name: Redis interview_transcript_line: - speaker_name: Turing transcript: Hello, my name is Tracy. This will be a short conversation... - speaker_name: Lav Shah transcript: I do have experience with front-end development as well. agent_response_summary: - question: text: How did you design the retry mechanism for failed messages? agents: - id: ai_researcher_1 name: Chris role: AI researcher total_weighted_score: 72 maximum_score: 90 criteria: retry_mechanism_design: score: 4 weight: 6 weighted: 24 maximum_score: 30 problemsolving_skills: score: 4 weight: 4 weighted: 16 maximum_score: 20 step6_summary: summary: >- The candidate provided a detailed explanation of the retry mechanism. retry_mechanism_design: >- They mentioned using two retry consumers after fifteen minutes. normalization: maximum_possible: 540 final_scores: ai_researcher_1: 72 ai_researcher_1_max: 90 ai_researcher_2: 72 ai_researcher_2_max: 90 ai_researcher_normalized: 0.8 ai_researcher_1_normalized: 0.8 mapping: agent: ai_researcher: AI researcher subject_matter_expert: Subject Matter Expert criteria: retry_mechanism_design: Retry Mechanism Design problemsolving_skills: Problem-Solving Skills assets: pdf_report_url: >- https://cdn.qode.gg/assessment-pdf-turing/93381e1c_assessment_profile.pdf video_url: https://cdn.qode.gg/recording/plDsMyRn7kWgRloewoDN/video_record.mp4 transcript_url: https://cdn.qode.gg/recording/plDsMyRn7kWgRloewoDN/chapter.json metadata: start_time: '2025-11-25T10:00:07.933Z' end_time: '2025-11-25T10:15:18.870Z' scorecard_time: '2025-11-25T10:16:00.000Z' scheduled_time: '2025-11-17T23:15:03.136Z' drop_off_time: '2025-11-17T23:15:03.136Z' realtime_status: INTERVIEW_COMPLETED is_terminal: true duration: 1911 interview_signals: tab_changes: - 192.173 abnormal_eye_tracking: [] face_out_of_view: - 184 - 808 avg_latency: 2 external_screens: false black_screen_timestamps: - start: '00:00:01.600' end: '00:00:03.000' duration: 1.4 - start: '00:01:42.711' end: '00:01:48.411' duration: 5.7 meeting_data: completed_status: COMPLETED start_date_time: '2026-03-10T03:02:01.438Z' end_date_time: '2026-03-10T03:02:01.438Z' chapters: - title: Explanation of the approach to the first task start_time: '00:00:17.218' end_time: '00:03:20.830' - title: Explanation of the approach to the second task start_time: '00:03:20.830' end_time: '00:05:18.888' - title: Key technical areas of expertise start_time: '00:05:18.888' end_time: null question_summary: - question: >- Could you walk me through your approach and explain how you intended to solve the problem? summary: They chose to skip this task. timestamp: '00:00:17.218' question_score: agent_summary: Not evaluated score: score: 0 ci_lower: 0 ci_upper: 0 skills_score: - name: Algorithm Design score: score: 0 ci_lower: 0 ci_upper: 0 summary: Not evaluated agents: - name: Jordan summary: '' - name: Jordan summary: '' - name: Programming score: score: 0 ci_lower: 0 ci_upper: 0 summary: Not evaluated agents: - name: Jordan summary: '' - name: Jordan summary: '' - question: >- Could you briefly describe your approach and the reasoning behind your implementation? summary: They chose to skip this question in the interview. timestamp: '00:03:20.830' question_score: agent_summary: Not evaluated score: score: 0 ci_lower: 0 ci_upper: 0 skills_score: - name: Algorithm Design score: score: 0 ci_lower: 0 ci_upper: 0 summary: Not evaluated agents: - name: Jordan summary: '' - name: Jordan summary: '' - name: Programming score: score: 0 ci_lower: 0 ci_upper: 0 summary: Not evaluated agents: - name: Jordan summary: '' - name: Jordan summary: '' - name: Data Structures score: score: 0 ci_lower: 0 ci_upper: 0 summary: Not evaluated agents: - name: Jordan summary: '' - name: Jordan summary: '' - question: >- Could you share the three to five key technical areas that best represent your expertise? summary: They expressed a desire to end the interview. timestamp: '00:05:18.888' question_score: null skills_score: [] raw_transcript: > Tracy at Qode: Hello, my name is Tracy. Just before we begin—how's your day been so far? thanh.le+coding35: Good. Good, Tracy. Tracy at Qode: I'm glad to hear that! Before we continue, please hold your ID card up to the camera so we can take a quick screenshot for verification. Thank you! Great! This will be a technical assessment to evaluate your skills. The assessment will be presented on your screen now. Thanks for confirming you're ready. Here's how the interview will proceed: - There are **2 coding tasks**. - Each task has a **15-minute time limit**. - After the coding tasks, we'll move on to some technical interview questions. - If you get stuck on a task, you can skip it at any time by letting me know. Let's begin. **Task 1 of 2: Special Permutations (Inversions)** You are given an array of the first m natural numbers... overall_score: 70 interview_result: Pass count_all_core_domains: 3 count_covered_core_domains: 3 uncovered_domains: - Data & DB - DevOps assessment_version: pf_domain_3p_v4 card_check_image_path: >- https://cdn.qode.gg/recording/qode-v3-44e69082-259f-47db-a136-fd92c5d16cc8/passed_1_0.59_card_detected_1772449871458.jpg demonstrated_skills_across_domains: - domain_name: Technical Knowledge domain_category: Core observations: 5 domain_weight: 0.4 ci_lower: 70 ci_upper: 75 stability: Highly Reliable mean_score_across_skills: 72 skills: - Frontend Development - Backend Development scored_skills: - skill_name: Frontend Development pooled_mean_score: 75 ci_lower: 73 ci_upper: 77 stability: Highly Reliable interview_level_bootstrap: ci_lower: 68 ci_upper: 72 label: Highly Reliable interpretation: The interview outcome is highly precise and reliable. grading_rubric_result: criteria: - name: 'Coding test: 2 tasks' score: 0 weight: 0.4 - name: Resume & Technical deep-dive interview with Tracy score: 0 weight: 0.6 language_proficiency_scores: coherence_score: 75 fluency_score: 75 grammar_score: 75 pronunciation_score: 75 prosody_score: 75 accuracy_score: 75 completeness_score: 75 summary: The candidate is proficient in English. vocabulary_score: 75 language: en code_task_result: tasks: - name: Task 1 language_name: Python (3.8.1) testcase_passrate: 80% (4/5) code_execution_time: 0.05 implementation_time: 68.4 ram_consumption: 4379 algorithm_complexity: O(n) status: Failed coding_score: 0 latency_summary: turn_details: - turn: 1 speaker: Tracy latency: 2.5 content: Hello, my name is Tracy. start_time: '2025-11-25T10:00:07.933Z' end_time: '2025-11-25T10:00:10.433Z' - turn: 2 speaker: Candidate latency: 1.8 content: Hi Tracy, nice to meet you. start_time: '2025-11-25T10:00:10.433Z' end_time: '2025-11-25T10:00:12.233Z' median: 2.3 avg_raw: 3.1 avg_clean: 2.4 anomaly_count: 1 anomaly_turns: - 3 event: REPORT_GENERATED timestamp: '2025-11-25T10:16:00.000Z' InterviewSignals: type: object nullable: true properties: tab_changes: type: array description: Timestamps (in seconds) where the candidate switched browser tabs. items: type: integer example: 12 default: [] example: - 12 - 40 - 120 abnormal_eye_tracking: type: array description: Timestamps (in seconds) where abnormal eye-tracking was detected. items: type: integer example: 45 default: [] example: - 12 - 140 - 220 face_out_of_view: type: array description: >- Timestamps (in seconds) where the candidate's face was out of camera view. items: type: integer example: 78 default: [] example: - 12 - 410 - 520 avg_latency: type: number description: Average audio/video latency during the interview (in seconds). example: 3 external_screens: type: boolean description: Whether external monitors/screens were detected during the session. example: false AgentResponseSummaryItem: type: object description: Evaluation of one candidate answering one interview question. properties: question: type: object properties: text: type: string description: Full question text presented to the candidate. example: How did you design the retry mechanism for failed messages? agents: type: array description: All evaluator personas and their scores for this question. items: $ref: '#/components/schemas/AgentEvaluation' normalization: $ref: '#/components/schemas/NormalizationObject' mapping: $ref: '#/components/schemas/MappingObject' LanguageProficiencyScoreItem: type: object description: >- Language proficiency scores for one language (base fields only; used in other_languages array). properties: coherence_score: type: number description: Coherence score for the assessed language. fluency_score: type: number description: Fluency score for the assessed language. grammar_score: type: number description: Grammar score for the assessed language. pronunciation_score: type: number description: Pronunciation score for the assessed language. prosody_score: type: number description: Prosody score for the assessed language. accuracy_score: type: number description: Accuracy score for the assessed language. completeness_score: type: number description: Completeness score for the assessed language. summary: type: string description: Summary of the language proficiency assessment. vocabulary_score: type: number description: Vocabulary score for the assessed language. language: type: string description: ISO or display code for the assessed language. required: - coherence_score - fluency_score - grammar_score - pronunciation_score - summary - vocabulary_score AgentEvaluation: type: object properties: id: type: string description: Index string of the agent, can use mapping object to map. example: ai_researcher_1 name: type: string description: Agent persona name. example: Chris role: type: string description: >- Evaluator persona type (e.g., AI Researcher, Subject Matter Expert, Project Manager). example: AI researcher total_weighted_score: type: number description: 'Final score for this agent for this question: Σ (score_k × weight_k)' example: 72 maximum_score: type: number description: 'Maximum possible score for this question: Σ (weight_k × 5)' example: 90 criteria: type: object description: Scores for each rubric criterion for this agent. additionalProperties: $ref: '#/components/schemas/CriterionScore' step6_summary: type: object description: Human-readable interpretation of the agent's evaluation. properties: summary: type: string description: >- High-level summary of the candidate's performance on this question. additionalProperties: type: string NormalizationObject: type: object description: Aggregates all agent-level scores for this question. properties: maximum_possible: type: number description: Total possible score across all agents for this question. example: 540 final_scores: type: object description: Each agent's actual weighted score and maximum possible score. additionalProperties: type: number additionalProperties: type: number MappingObject: type: object description: Translation layer between internal keys and human-readable labels. properties: agent: type: object description: Maps internal agent-role identifiers to human-readable labels. additionalProperties: type: string example: ai_researcher: AI researcher subject_matter_expert: Subject Matter Expert project_manager: Project Manager criteria: type: object description: Maps internal criterion keys to their display names. additionalProperties: type: string example: retry_mechanism_design: Retry Mechanism Design problemsolving_skills: Problem-Solving Skills CriterionScore: type: object properties: score: type: number description: >- Agent's rating on a 0–5 scale (0 = no evidence, 5 = strongest demonstration). example: 4 weight: type: number description: Relative importance of the criterion in the rubric. example: 6 weighted: type: number description: 'Computed as: score × weight' example: 24 maximum_score: type: number description: 'Computed as: 5 × weight' example: 30 securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: JWT ````