app: description: '' icon: 🤖 icon_background: '#FFEAD5' mode: workflow name: 'OpenAI Audio Toolkit Demo: Step-by-Step Diarization' use_icon_as_answer_icon: false dependencies: - current_identifier: null type: package value: plugin_unique_identifier: kurokobo/openai_audio_toolkit:0.0.1@dbc3cdff903953ac93b7f6e0315ec02a867d747fd3da8586337a53cb6559863f version: null kind: app version: 0.5.0 workflow: conversation_variables: [] environment_variables: [] features: file_upload: allowed_file_extensions: - .JPG - .JPEG - .PNG - .GIF - .WEBP - .SVG allowed_file_types: - image allowed_file_upload_methods: - local_file - remote_url enabled: false fileUploadConfig: attachment_image_file_size_limit: 2 audio_file_size_limit: 1024 batch_count_limit: 5 file_size_limit: 15 file_upload_limit: 20 image_file_batch_limit: 10 image_file_size_limit: 10 single_chunk_attachment_limit: 10 video_file_size_limit: 1024 workflow_file_upload_limit: 10 image: enabled: false number_limits: 3 transfer_methods: - local_file - remote_url number_limits: 3 opening_statement: '' retriever_resource: enabled: true sensitive_word_avoidance: enabled: false speech_to_text: enabled: false suggested_questions: [] suggested_questions_after_answer: enabled: false text_to_speech: enabled: false language: '' voice: '' graph: edges: - data: isInIteration: false isInLoop: false sourceType: start targetType: iteration id: 1770471165089-source-1770474034524-target source: '1770471165089' sourceHandle: source target: '1770474034524' targetHandle: target type: custom zIndex: 0 - data: isInIteration: true isInLoop: false iteration_id: '1770474034524' sourceType: iteration-start targetType: tool id: 1770474034524start-source-1770474044550-target source: 1770474034524start sourceHandle: source target: '1770474044550' targetHandle: target type: custom zIndex: 1002 - data: isInIteration: false isInLoop: false sourceType: iteration targetType: iteration id: 1770474034524-source-1770474050573-target source: '1770474034524' sourceHandle: source target: '1770474050573' targetHandle: target type: custom zIndex: 0 - data: isInIteration: true isInLoop: false iteration_id: '1770474050573' sourceType: iteration-start targetType: tool id: 1770474050573start-source-1770474058561-target source: 1770474050573start sourceHandle: source target: '1770474058561' targetHandle: target type: custom zIndex: 1002 - data: isInIteration: false isInLoop: false sourceType: iteration targetType: tool id: 1770474050573-source-1770474065919-target source: '1770474050573' sourceHandle: source target: '1770474065919' targetHandle: target type: custom zIndex: 0 - data: isInIteration: false isInLoop: false sourceType: tool targetType: tool id: 1770474065919-source-1770474083718-target source: '1770474065919' sourceHandle: source target: '1770474083718' targetHandle: target type: custom zIndex: 0 - data: isInIteration: false isInLoop: false sourceType: tool targetType: end id: 1770474083718-source-1770474225480-target source: '1770474083718' sourceHandle: source target: '1770474225480' targetHandle: target type: custom zIndex: 0 nodes: - data: selected: false title: User Input type: start variables: - allowed_file_extensions: [] allowed_file_types: - audio - video allowed_file_upload_methods: - local_file - remote_url default: '' hint: '' label: Audio or Video Files to be Diarized max_length: 10 options: [] placeholder: '' required: true type: file-list variable: input_files height: 109 id: '1770471165089' position: x: 80 y: 282 positionAbsolute: x: 80 y: 282 selected: false sourcePosition: right targetPosition: left type: custom width: 242 - data: error_handle_mode: terminated flatten_output: true height: 202 is_parallel: false iterator_input_type: array[file] iterator_selector: - '1770471165089' - input_files output_selector: - '1770474044550' - files output_type: array[file] parallel_nums: 10 selected: false start_node_id: 1770474034524start title: Iteration type: iteration width: 386 height: 202 id: '1770474034524' position: x: 383 y: 282 positionAbsolute: x: 383 y: 282 selected: false sourcePosition: right targetPosition: left type: custom width: 386 zIndex: 1 - data: desc: '' isInIteration: true selected: false title: '' type: iteration-start draggable: false height: 48 id: 1770474034524start parentId: '1770474034524' position: x: 24 y: 68 positionAbsolute: x: 407 y: 350 selectable: false sourcePosition: right targetPosition: left type: custom-iteration-start width: 44 zIndex: 1002 - data: isInIteration: true isInLoop: false is_team_authorization: true iteration_id: '1770474034524' paramSchemas: - auto_generate: null default: null form: form human_description: en_US: One or more audio/video files to process; split if necessary based on size and duration limits. ja_JP: 処理する 1 つ以上の音声・動画ファイル。サイズと長さの制限に基づいて必要に応じて分割されます。 pt_BR: Um ou mais arquivos de áudio/vídeo para processar; divididos conforme necessário com base nos limites de tamanho e duração. zh_Hans: 一个或多个音频/视频文件要处理；如果必要，根据大小和持续时间限制进行拆分。 label: en_US: Audio/Video Files ja_JP: 音声・動画ファイル（複数可） pt_BR: Arquivos de áudio/vídeo zh_Hans: 音频/视频文件（可多个） llm_description: Provide one or more audio/video files to be split if necessary based on size (25MB) and duration (1500s) limits. max: null min: null name: input_files options: [] placeholder: null precision: null required: true scope: null template: null type: files - auto_generate: null default: false form: form human_description: en_US: When splitting is needed, split audio at silence points instead of fixed time intervals. This may be slower but produces more natural splits. Falls back to time-based splitting if no silence is detected. ja_JP: 分割が必要な場合、固定時間ではなく無音部分で音声を分割します。処理は遅くなりますが、より自然な分割ができます。無音が検出されない場合は時間ベースの分割にフォールバックします。 pt_BR: Quando a divisão é necessária, divide o áudio em pontos de silêncio em vez de intervalos de tempo fixos. Pode ser mais lento, mas produz divisões mais naturais. Volta à divisão baseada em tempo se nenhum silêncio for detectado. zh_Hans: 当需要拆分时，在静音点而非固定时间间隔处拆分音频。这可能较慢，但会产生更自然的拆分。如果未检测到静音，将回退到基于时间的拆分。 label: en_US: Use Silence Detection ja_JP: 無音検出を使用 pt_BR: Usar detecção de silêncio zh_Hans: 使用静音检测 llm_description: Split audio at detected silence points for more natural chunks when splitting is needed. Defaults to time-based splitting; slower processing. max: null min: null name: use_silence_detection options: [] placeholder: null precision: null required: false scope: null template: null type: boolean params: input_files: '' use_silence_detection: '' plugin_id: kurokobo/openai_audio_toolkit plugin_unique_identifier: kurokobo/openai_audio_toolkit:0.0.1@2aa30cb18cb7cff76cd4026c256f947fe6894d32163702e64770779e3c23bd1e provider_icon: /console/api/workspaces/current/plugin/icon?tenant_id=5aa2823f-93af-4737-9833-51d5b2a70f20&filename=e7422329b9a0aa8af892e5cb521c6fa02cf1217fe446432a3ca1a2ee74408dcc.svg provider_id: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_name: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_type: builtin selected: true title: Split Audio tool_configurations: input_files: type: variable value: - '1770474034524' - item use_silence_detection: type: constant value: false tool_description: Splits one or more audio/video files based on file size and duration limits. API-native formats within 25MB and 1500 seconds are passed through; other inputs are transcoded and/or split into smaller chunks. MP4 with supported audio codecs may be extracted as audio. tool_label: Split Audio tool_name: split_audio tool_node_version: '2' tool_parameters: {} type: tool height: 114 id: '1770474044550' parentId: '1770474034524' position: x: 128 y: 68 positionAbsolute: x: 511 y: 350 selected: true sourcePosition: right targetPosition: left type: custom width: 242 zIndex: 1002 - data: error_handle_mode: terminated flatten_output: true height: 176 is_parallel: false iterator_input_type: array[file] iterator_selector: - '1770474034524' - output output_selector: - '1770474058561' - text output_type: array[string] parallel_nums: 10 selected: false start_node_id: 1770474050573start title: Iteration 2 type: iteration width: 386 height: 176 id: '1770474050573' position: x: 823 y: 282 positionAbsolute: x: 823 y: 282 selected: false sourcePosition: right targetPosition: left type: custom width: 386 zIndex: 1 - data: desc: '' isInIteration: true selected: false title: '' type: iteration-start draggable: false height: 48 id: 1770474050573start parentId: '1770474050573' position: x: 24 y: 68 positionAbsolute: x: 847 y: 350 selectable: false sourcePosition: right targetPosition: left type: custom-iteration-start width: 44 zIndex: 1002 - data: isInIteration: true isInLoop: false is_team_authorization: true iteration_id: '1770474050573' paramSchemas: - auto_generate: null default: null form: form human_description: en_US: One or more audio files to transcribe with speaker diarization enabled; processed in the given order. Inputs must already be accepted by the API (e.g., split_audio outputs). ja_JP: '話者分離を有効にして文字起こしする 1 つ以上の音声ファイル。指定順に結合されます。入力は API が受け付ける形式である前提です（例: split_audio の出力）。' pt_BR: 'Um ou mais arquivos de áudio para transcrição com diarização de falantes; serão concatenados na ordem fornecida. As entradas devem ser aceitas pela API (ex.: saída do split_audio).' zh_Hans: 一个或多个启用说话人分离进行转录的音频文件，将按提供顺序合并。输入必须已被 API 接受（例如 split_audio 的输出）。 label: en_US: Audio Files ja_JP: 音声ファイル（複数可） pt_BR: Arquivos de áudio zh_Hans: 音频文件（可多个） llm_description: Provide one or more audio files to transcribe with diarization; results are concatenated in order. Inputs must already be accepted by the API. max: null min: null name: input_files options: [] placeholder: null precision: null required: true scope: null template: null type: files params: input_files: '' plugin_id: kurokobo/openai_audio_toolkit plugin_unique_identifier: kurokobo/openai_audio_toolkit:0.0.1@2aa30cb18cb7cff76cd4026c256f947fe6894d32163702e64770779e3c23bd1e provider_icon: /console/api/workspaces/current/plugin/icon?tenant_id=5aa2823f-93af-4737-9833-51d5b2a70f20&filename=e7422329b9a0aa8af892e5cb521c6fa02cf1217fe446432a3ca1a2ee74408dcc.svg provider_id: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_name: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_type: builtin selected: false title: Diarize Audio tool_configurations: input_files: type: variable value: - '1770474050573' - item tool_description: Transcribes one or more audio files with speaker diarization and outputs the concatenated segments as text and JSON. Inputs must already be accepted by the API (e.g., outputs from split_audio). tool_label: Diarize Audio tool_name: diarize_audio tool_node_version: '2' tool_parameters: {} type: tool height: 88 id: '1770474058561' parentId: '1770474050573' position: x: 128 y: 68 positionAbsolute: x: 951 y: 350 selected: false sourcePosition: right targetPosition: left type: custom width: 242 zIndex: 1002 - data: is_team_authorization: true paramSchemas: - auto_generate: null default: null form: form human_description: en_US: JSON string of an array of objects that each contain a segments array and optional metadata. Each item should match diarize_audio output shape. ja_JP: segments 配列と任意の metadata を持つオブジェクト配列の JSON 文字列。各要素は diarize_audio の出力形状に合わせます。 pt_BR: String JSON de um array de objetos contendo um array segments e metadata opcional. Cada item deve seguir o formato do diarize_audio. zh_Hans: 包含 segments 数组和可选 metadata 的对象数组的 JSON 字符串。每个元素应符合 diarize_audio 的输出结构。 label: en_US: Items (JSON String) ja_JP: アイテム（JSON文字列） pt_BR: Itens (JSON String) zh_Hans: 项目（JSON 字符串） llm_description: Provide a JSON string of diarize_audio-like outputs. Each item must include segments and may include metadata. max: null min: null name: items_json_string options: [] placeholder: null precision: null required: false scope: null template: null type: string - auto_generate: null default: null form: form human_description: en_US: (experimental) Array of objects that each contain a segments array and optional metadata. Each item should match diarize_audio output shape. ja_JP: (experimental) segments 配列と任意の metadata を持つオブジェクトの配列。各要素は diarize_audio の出力形状に合わせます。 pt_BR: (experimental) Array de objetos contendo um array segments e metadata opcional. Cada item deve seguir o formato do diarize_audio. zh_Hans: (experimental) 包含 segments 数组和可选 metadata 的对象数组。每个元素应符合 diarize_audio 的输出结构。 label: en_US: Items (Array, experimental) ja_JP: アイテム（配列・experimental） pt_BR: Itens (Array, experimental) zh_Hans: 项目（数组・experimental） llm_description: (experimental) Provide an array of diarize_audio-like outputs. Each item must include segments and may include metadata. max: null min: null name: items_array options: [] placeholder: null precision: null required: false scope: null template: null type: array params: items_array: '' items_json_string: '' plugin_id: kurokobo/openai_audio_toolkit plugin_unique_identifier: kurokobo/openai_audio_toolkit:0.0.1@2aa30cb18cb7cff76cd4026c256f947fe6894d32163702e64770779e3c23bd1e provider_icon: /console/api/workspaces/current/plugin/icon?tenant_id=5aa2823f-93af-4737-9833-51d5b2a70f20&filename=e7422329b9a0aa8af892e5cb521c6fa02cf1217fe446432a3ca1a2ee74408dcc.svg provider_id: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_name: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_type: builtin selected: false title: Concat Segments tool_configurations: items_array: type: variable value: - '1770474050573' - output items_json_string: type: mixed value: null tool_description: Concatenates diarization segment arrays, normalizes segment IDs and time offsets, and outputs the result as text and JSON. tool_label: Concat Segments tool_name: concat_segments tool_node_version: '2' tool_parameters: {} type: tool height: 114 id: '1770474065919' position: x: 383 y: 535 positionAbsolute: x: 383 y: 535 selected: false sourcePosition: right targetPosition: left type: custom width: 242 - data: is_team_authorization: true paramSchemas: - auto_generate: null default: null form: form human_description: en_US: JSON string of an object or array that contains segments (e.g., diarize_audio / concat_segments / replace_speaker_name text output). ja_JP: 'segments を含むオブジェクトまたは配列の JSON 文字列（例: diarize_audio / concat_segments / replace_speaker_name の text 出力）。' pt_BR: 'String JSON de um objeto ou array contendo segments (ex.: saída text do diarize_audio / concat_segments / replace_speaker_name).' zh_Hans: 包含 segments 的对象或数组的 JSON 字符串（例如 diarize_audio / concat_segments / replace_speaker_name 的 text 输出）。 label: en_US: Segments (JSON String) ja_JP: セグメント（JSON文字列） pt_BR: Segmentos (JSON String) zh_Hans: 分段（JSON 字符串） llm_description: Provide a JSON string of a diarize_audio/concat_segments-like output. max: null min: null name: segments_json_string options: [] placeholder: null precision: null required: true scope: null template: null type: string - auto_generate: null default: plain_text form: form human_description: en_US: Choose the output format and delivery method for the formatted transcript. ja_JP: 整形された書き起こしの出力形式と出力方法を選択します。 pt_BR: Escolha o formato e o modo de saída do transcript formatado. zh_Hans: 选择格式化转录的输出格式和输出方式。 label: en_US: Output Format ja_JP: 出力フォーマット pt_BR: Formato de saída zh_Hans: 输出格式 llm_description: Choose output format and mode, e.g., plain_text or vtt_file. max: null min: null name: output_format options: - icon: '' label: en_US: Plain (Text) ja_JP: Plain（テキスト） pt_BR: Plain (Texto) zh_Hans: Plain（文本） value: plain_text - icon: '' label: en_US: JSON (Text) ja_JP: JSON（テキスト） pt_BR: JSON (Texto) zh_Hans: JSON（文本） value: json_text - icon: '' label: en_US: Markdown (Text) ja_JP: Markdown（テキスト） pt_BR: Markdown (Texto) zh_Hans: Markdown（文本） value: markdown_text - icon: '' label: en_US: VTT (Text) ja_JP: VTT（テキスト） pt_BR: VTT (Texto) zh_Hans: VTT（文本） value: vtt_text - icon: '' label: en_US: SRT (Text) ja_JP: SRT（テキスト） pt_BR: SRT (Texto) zh_Hans: SRT（文本） value: srt_text - icon: '' label: en_US: Plain (File) ja_JP: Plain（ファイル） pt_BR: Plain (Arquivo) zh_Hans: Plain（文件） value: plain_file - icon: '' label: en_US: JSON (File) ja_JP: JSON（ファイル） pt_BR: JSON (Arquivo) zh_Hans: JSON（文件） value: json_file - icon: '' label: en_US: Markdown (File) ja_JP: Markdown（ファイル） pt_BR: Markdown (Arquivo) zh_Hans: Markdown（文件） value: markdown_file - icon: '' label: en_US: VTT (File) ja_JP: VTT（ファイル） pt_BR: VTT (Arquivo) zh_Hans: VTT（文件） value: vtt_file - icon: '' label: en_US: SRT (File) ja_JP: SRT（ファイル） pt_BR: SRT (Arquivo) zh_Hans: SRT（文件） value: srt_file placeholder: null precision: null required: false scope: null template: null type: select params: output_format: '' segments_json_string: '' plugin_id: kurokobo/openai_audio_toolkit plugin_unique_identifier: kurokobo/openai_audio_toolkit:0.0.1@2aa30cb18cb7cff76cd4026c256f947fe6894d32163702e64770779e3c23bd1e provider_icon: /console/api/workspaces/current/plugin/icon?tenant_id=5aa2823f-93af-4737-9833-51d5b2a70f20&filename=e7422329b9a0aa8af892e5cb521c6fa02cf1217fe446432a3ca1a2ee74408dcc.svg provider_id: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_name: kurokobo/openai_audio_toolkit/openai_audio_toolkit provider_type: builtin selected: false title: Format Segments tool_configurations: output_format: type: constant value: markdown_text segments_json_string: type: mixed value: '{{#1770474065919.text#}}' tool_description: Formats diarization segments into text, Markdown, VTT, or SRT. tool_label: Format Segments tool_name: format_segments tool_node_version: '2' tool_parameters: {} type: tool height: 114 id: '1770474083718' position: x: 680.2 y: 535 positionAbsolute: x: 680.2 y: 535 selected: false sourcePosition: right targetPosition: left type: custom width: 242 - data: outputs: - value_selector: - '1770474083718' - text value_type: string variable: text selected: false title: Output type: end height: 88 id: '1770474225480' position: x: 971.2 y: 535 positionAbsolute: x: 971.2 y: 535 selected: false sourcePosition: right targetPosition: left type: custom width: 242 viewport: x: -315.20000000000005 y: -174 zoom: 1 rag_pipeline_variables: []