> ## Documentation Index > Fetch the complete documentation index at: https://docs.mixpeek.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # List Collections > This endpoint allows you to list collections. ## OpenAPI ````yaml post /v1/collections/list openapi: 3.1.0 info: title: Mixpeek API description: >- This is the Mixpeek API, providing access to various endpoints for data processing and retrieval. termsOfService: https://mixpeek.com/terms contact: name: Mixpeek Support url: https://mixpeek.com/contact email: info@mixpeek.com version: '0.82' servers: - url: https://api.mixpeek.com description: Production security: [] paths: /v1/collections/list: post: tags: - Collections summary: List Collections description: This endpoint allows you to list collections. operationId: list_collections_v1_collections_list_post parameters: - name: limit in: query required: false schema: anyOf: - type: integer maximum: 1000 minimum: 1 - type: 'null' title: Limit - name: page_size in: query required: false schema: anyOf: - type: integer maximum: 1000 minimum: 1 - type: 'null' title: Page Size - name: offset in: query required: false schema: anyOf: - type: integer maximum: 10000 minimum: 0 - type: 'null' title: Offset - name: page in: query required: false schema: anyOf: - type: integer minimum: 1 - type: 'null' title: Page - name: cursor in: query required: false schema: anyOf: - type: string - type: 'null' title: Cursor - name: include_total in: query required: false schema: type: boolean default: false title: Include Total requestBody: content: application/json: schema: $ref: '#/components/schemas/ListCollectionsRequest' responses: '200': description: Successful Response content: application/json: schema: $ref: '#/components/schemas/ListCollectionsResponse' '400': description: Bad Request content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '401': description: Unauthorized content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '403': description: Forbidden content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '404': description: Not Found content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '422': description: Validation Error content: application/json: schema: $ref: '#/components/schemas/HTTPValidationError' '500': description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' components: schemas: ListCollectionsRequest: properties: filters: anyOf: - additionalProperties: true type: object - type: 'null' title: Filters description: >- Filters to apply when listing collections. Supports nested field filtering like 'taxonomy_applications.taxonomy_id'. Format: {"AND": [{"field": "field_name", "operator": "eq", "value": "value"}]} sort: anyOf: - $ref: '#/components/schemas/SortOption' - type: 'null' description: Sort options for the results search: anyOf: - type: string - type: 'null' title: Search description: >- Search term for wildcard search across collection_id, collection_name, description, and other text fields case_sensitive: type: boolean title: Case Sensitive description: If True, filters and search will be case-sensitive default: false type: object title: ListCollectionsRequest description: >- Request model for listing collections. To filter by taxonomy, use dot notation in filters: filters.AND = [{"field": "taxonomy_applications.taxonomy_id", "operator": "eq", "value": "tax_123"}] ListCollectionsResponse: properties: results: items: $ref: '#/components/schemas/CollectionResponse' type: array title: Results description: List of collections pagination: $ref: '#/components/schemas/PaginationResponse' description: Pagination information total_count: type: integer title: Total Count description: Total number of collections matching the query stats: anyOf: - $ref: '#/components/schemas/CollectionListStats' - type: 'null' description: Aggregate statistics across all collections in the result type: object required: - results - pagination - total_count title: ListCollectionsResponse description: Response model for listing collections. ErrorResponse: properties: success: type: boolean title: Success description: Always false for error responses default: false status: type: integer title: Status description: HTTP status code for this error error: $ref: '#/components/schemas/ErrorDetail' description: Error details payload type: object required: - status - error title: ErrorResponse description: Error response model. examples: - error: details: id: ns_123 resource: namespace message: Namespace not found type: NotFoundError status: 404 success: false HTTPValidationError: properties: detail: items: $ref: '#/components/schemas/ValidationError' type: array title: Detail type: object title: HTTPValidationError SortOption: properties: field: type: string title: Field description: Field to sort by, supports dot notation for nested fields example: created_at direction: $ref: '#/components/schemas/SortDirection' description: Sort direction default: asc example: desc type: object required: - field title: SortOption description: |- Specifies how to sort query results. Attributes: field: Field to sort by direction: Sort direction (ascending or descending) CollectionResponse: properties: collection_id: type: string title: Collection Id description: >- NOT REQUIRED (auto-generated). Unique identifier for this collection. Used for: API paths, document queries, pipeline references. Format: 'col_' prefix + 10 random alphanumeric characters. Stable after creation - use for all collection references. examples: - col_a1b2c3d4e5 - col_xyz789abc - col_video_frames collection_name: type: string maxLength: 100 minLength: 3 title: Collection Name description: >- REQUIRED. Human-readable name for the collection. Must be unique within the namespace. Used for: Display, lookups (can query by name or ID), organization. Format: Alphanumeric with underscores/hyphens, 3-100 characters. Examples: 'product_embeddings', 'video_frames', 'customer_documents'. examples: - video_frames - product_embeddings - customer_documents description: anyOf: - type: string - type: 'null' title: Description description: >- NOT REQUIRED. Human-readable description of the collection's purpose. Use for: Documentation, team communication, UI display. Common pattern: Describe what the collection contains and what processing is applied. examples: - Video frames extracted at 1 FPS with CLIP embeddings - Product catalog with image embeddings and metadata input_schema: anyOf: - $ref: '#/components/schemas/BucketSchema-Output' - type: 'null' description: >- Auto-computed from source. Input schema defining fields available to the feature extractor. Source: bucket.bucket_schema (if source.type='bucket') OR upstream_collection.output_schema (if source.type='collection'). Determines: Which fields can be used in input_mappings and field_passthrough. This is the 'left side' of the transformation - what data goes IN. Format: BucketSchema with properties dict. Use for: Validating input_mappings, configuring field_passthrough. examples: - properties: title: type: string content: type: text - properties: video: type: video campaign_id: type: string output_schema: anyOf: - $ref: '#/components/schemas/BucketSchema-Output' - type: 'null' description: >- Auto-computed at creation. Output schema defining fields in final documents. Computed as: field_passthrough fields + extractor output fields (deterministic). Known IMMEDIATELY when collection is created - no waiting for documents! This is the 'right side' of the transformation - what data comes OUT. Use for: Understanding document structure, building queries, schema validation. Example: {title (passthrough), embedding (extractor output)} = output_schema. examples: - properties: title: type: string text_extractor_v1_embedding: type: array - properties: campaign_id: type: string multimodal_extractor_v1_embedding: type: array feature_extractor: $ref: >- #/components/schemas/shared__collection__features__extractors__models__FeatureExtractorConfig-Output description: >- REQUIRED. Single feature extractor configuration for this collection. Defines: extractor name/version, input_mappings, field_passthrough, parameters. Task 9 change: ONE extractor per collection (previously supported multiple). For multiple extractors: Create multiple collections and use collection-to-collection pipelines. Use field_passthrough to include additional source fields beyond extractor outputs. examples: - feature_extractor_name: text_extractor field_passthrough: - source_path: title input_mappings: text: content version: v1 source: $ref: '#/components/schemas/SourceConfig-Output' description: >- REQUIRED. Source configuration defining where data comes from. Type 'bucket': Process objects from one or more buckets (tier 1). Type 'collection': Process documents from upstream collection(s) (tier 2+). For multi-bucket sources, all buckets must have compatible schemas. For multi-collection sources, all collections must have compatible schemas. Determines input_schema and enables decomposition trees. examples: - bucket_ids: - bkt_articles type: bucket - bucket_ids: - bkt_us_products - bkt_eu_products type: bucket - collection_id: col_video_frames type: collection - collection_ids: - col_products_2023 - col_products_2024 type: collection source_bucket_schemas: anyOf: - additionalProperties: $ref: '#/components/schemas/BucketSchema-Output' type: object - type: 'null' title: Source Bucket Schemas description: >- NOT REQUIRED (auto-computed). Snapshot of bucket schemas at collection creation. Only populated for multi-bucket collections (source.type='bucket' with multiple bucket_ids). Key: bucket_id, Value: BucketSchema at time of collection creation. Used for: Schema compatibility validation, document lineage, debugging. Schema snapshot is immutable - bucket schema changes after collection creation do not affect this. Single-bucket collections may omit this field (schema in input_schema is sufficient). examples: - null - bkt_eu_products: properties: image: type: image title: type: string price: type: number required: - image - title bkt_us_products: properties: image: type: image title: type: string price: type: number required: - image - title source_lineage: items: $ref: '#/components/schemas/SingleLineageEntry' type: array title: Source Lineage description: >- NOT REQUIRED (auto-computed). Lineage chain showing complete processing history. Each entry contains: source_config, feature_extractor, output_schema for one tier. Length indicates processing depth (1 = tier 1, 2 = tier 2, etc.). Use for: Understanding multi-tier pipelines, visualizing decomposition trees. examples: - [] - - output_schema: {} source_config: bucket_id: bkt_videos type: bucket vector_indexes: items: {} type: array title: Vector Indexes description: >- NOT REQUIRED (auto-computed from extractor). Vector indexes for semantic search. Populated from feature_extractor.required_vector_indexes. Defines: Which embeddings are indexed, dimensions, distance metrics. Use for: Understanding search capabilities, debugging vector queries. examples: - [] - - distance: cosine name: text_extractor_v1_embedding size: 1024 payload_indexes: items: {} type: array title: Payload Indexes description: >- NOT REQUIRED (auto-computed from extractor + namespace). Payload indexes for filtering. Enables efficient filtering on metadata fields, timestamps, IDs. Populated from: extractor requirements + namespace defaults. Use for: Understanding which fields support fast filtering. examples: - [] - - field_name: collection_id type: keyword embedding_task: anyOf: - type: string - type: 'null' title: Embedding Task description: >- Override the embedding task hint for instruction-aware models (E5, Gemini). Defaults to 'retrieval_document' for indexing pipelines. Values: retrieval_document, retrieval_query, semantic_similarity, classification, clustering. Applied to all task-aware embedding models in this collection's extractor pipeline. enabled: type: boolean title: Enabled description: >- NOT REQUIRED (defaults to True). Whether the collection accepts new documents. False: Collection exists but won't process new objects. True: Active and processing. Use for: Temporarily disabling collections without deletion. default: true examples: - true - false metadata: anyOf: - additionalProperties: true type: object - type: 'null' title: Metadata description: >- NOT REQUIRED. Additional user-defined metadata for the collection. Arbitrary key-value pairs for custom organization, tracking, configuration. Not used by the platform - purely for user purposes. Common uses: team ownership, project tags, deployment environment. examples: - environment: production project: Q4_campaign team: data-science - null schedule: anyOf: - $ref: '#/components/schemas/CollectionScheduleConfig' - type: 'null' description: >- NOT REQUIRED. Schedule configuration for automatic re-processing. When set, a COLLECTION_TRIGGER trigger is created and linked to this collection. trigger_id: anyOf: - type: string - type: 'null' title: Trigger Id description: >- NOT REQUIRED. ID of the linked trigger for scheduled re-processing. Automatically set when a schedule is configured. created_at: anyOf: - type: string format: date-time - type: 'null' title: Created At description: >- Timestamp when the collection was created. Automatically set by the system when the collection is first saved to the database. updated_at: anyOf: - type: string format: date-time - type: 'null' title: Updated At description: >- Timestamp when the collection was last updated. Automatically updated by the system whenever the collection is modified. lifecycle_state: type: string title: Lifecycle State description: >- Storage lifecycle state: 'active' (Qdrant + S3), 'cold' (S3 only), 'archived' (metadata only). Managed via lifecycle API. default: active s3_vector_index: anyOf: - type: string - type: 'null' title: S3 Vector Index description: >- S3 Vectors index name for this collection (e.g. 'col_{collection_id}'). last_lifecycle_transition: anyOf: - type: string format: date-time - type: 'null' title: Last Lifecycle Transition description: Timestamp of the most recent lifecycle state change. tiering_rules: anyOf: - items: $ref: '#/components/schemas/TieringRule' type: array - type: 'null' title: Tiering Rules description: Automatic storage tiering rules (not enforced in V1). document_count: anyOf: - type: integer - type: 'null' title: Document Count description: Number of documents in the collection schema_version: type: integer title: Schema Version description: >- Version number for the output_schema. Increments automatically when schema is updated via document sampling. Used to track schema evolution and trigger downstream collection schema updates. default: 1 last_schema_sync: anyOf: - type: string format: date-time - type: 'null' title: Last Schema Sync description: >- Timestamp of last automatic schema sync from document sampling. Used to debounce schema updates (prevents thrashing). schema_sync_enabled: type: boolean title: Schema Sync Enabled description: >- Whether automatic schema discovery and sync is enabled for this collection. When True, schema is periodically updated by sampling documents. When False, schema remains fixed at creation time. default: true document_schema: anyOf: - additionalProperties: true type: object - type: 'null' title: Document Schema description: >- NOT REQUIRED. JSON Schema for validating documents on create/update. When set with schema_validation='strict', non-conforming documents are rejected (422). When set with schema_validation='warn', violations are recorded but document is accepted. Schema follows JSON Schema draft-07 format. Only validates user-defined fields (system fields like _internal, collection_id, document_id are excluded from validation). examples: - null - additionalProperties: true properties: title: maxLength: 200 type: string price: minimum: 0 type: number status: enum: - active - draft - archived type: string required: - title - price type: object schema_validation: type: string enum: - strict - warn - 'off' title: Schema Validation description: >- Document schema validation mode. 'strict': reject non-conforming documents with 422. 'warn': accept but attach _schema_violations field to the document. 'off': no validation (default, preserves current behavior). default: 'off' examples: - 'off' - strict - warn taxonomy_applications: anyOf: - items: $ref: '#/components/schemas/TaxonomyApplicationConfig-Output' type: array - type: 'null' title: Taxonomy Applications description: >- NOT REQUIRED. List of taxonomies to apply to documents in this collection. Each entry specifies: taxonomy_id, optional target_collection_id, optional filters. Enrichments are materialized (persisted to documents) during ingestion. Empty/null if no taxonomies attached. Use for: Categorization, hierarchical classification. examples: - null - - execution_mode: materialize taxonomy_id: tax_categories - execution_mode: materialize taxonomy_id: tax_brands cluster_applications: anyOf: - items: $ref: '#/components/schemas/ClusterApplicationConfig' type: array - type: 'null' title: Cluster Applications description: >- NOT REQUIRED. List of clusters to automatically execute when batch processing completes. Each entry specifies: cluster_id, auto_execute_on_batch, min_document_threshold, cooldown_seconds. Clusters enrich source documents with cluster assignments (cluster_id, cluster_label, etc.). Empty/null if no clusters attached. Use for: Segmentation, grouping, pattern discovery. examples: - null - - auto_execute_on_batch: true cluster_id: clust_product_categories cooldown_seconds: 3600 min_document_threshold: 100 alert_applications: anyOf: - items: $ref: '#/components/schemas/AlertApplicationConfig-Output' type: array - type: 'null' title: Alert Applications description: >- NOT REQUIRED. List of alerts to automatically execute when documents are ingested. Each entry specifies: alert_id, execution_mode, input_mappings, execution_phase, priority. Alerts run retrievers on ingested documents and send notifications when matches are found. Empty/null if no alerts attached. Use for: Content monitoring, safety detection, compliance alerts. examples: - null - - alert_id: alt_safety_001 execution_mode: on_ingest execution_phase: 3 input_mappings: - input_key: query_embedding source: path: features.video_embedding source_type: document_field priority: 100 retriever_enrichments: anyOf: - items: $ref: '#/components/schemas/RetrieverEnrichmentConfig-Output' type: array - type: 'null' title: Retriever Enrichments description: >- NOT REQUIRED. List of retriever enrichments to run on documents during post-processing. Each entry specifies: retriever_id, input_mappings, write_back_fields, execution_phase, priority. Retriever enrichments execute a retriever pipeline and write selected result fields back to documents. Empty/null if no enrichments attached. Use for: LLM classification, cross-collection joins, enrichment. examples: - null - - execution_phase: 4 input_mappings: - input_key: query source: path: title source_type: document_field priority: 0 retriever_id: ret_classifier_001 write_back_fields: - mode: first source_field: category target_field: _enrichment_category vector_count: anyOf: - type: integer - type: 'null' title: Vector Count description: >- Total number of vector entries across all documents in this collection. Computed as document_count * number_of_vector_indexes. Each document stores one vector per configured vector index (e.g. a collection with 1 embedding produces 1 vector per document). taxonomy_count: anyOf: - type: integer - type: 'null' title: Taxonomy Count description: Number of taxonomies connected to this collection retriever_count: anyOf: - type: integer - type: 'null' title: Retriever Count description: Number of retrievers connected to this collection type: object required: - collection_name - feature_extractor - source title: CollectionResponse description: Response model for collection endpoints. examples: - collection_id: col_a1b2c3d4e5 collection_name: article_embeddings description: >- Simple text collection: News articles with text embeddings from bucket source enabled: true feature_extractor: feature_extractor_name: text_extractor field_passthrough: - source_path: title input_mappings: text: content version: v1 input_schema: properties: title: type: string content: type: text output_schema: properties: title: type: string text_extractor_v1_embedding: type: array source: bucket_id: bkt_articles type: bucket - collection_id: col_xyz789abc collection_name: video_frames description: >- Video frames: Extracted at 1 FPS with CLIP embeddings and campaign_id passthrough enabled: true feature_extractor: feature_extractor_name: multimodal_extractor field_passthrough: - source_path: campaign_id input_mappings: video: video parameters: fps: 1 version: v1 input_schema: properties: video: type: video campaign_id: type: string output_schema: properties: campaign_id: type: string multimodal_extractor_v1_embedding: type: array source: bucket_id: bkt_marketing_videos type: bucket - collection_id: col_scenes_def collection_name: detected_scenes description: >- Tier 2 decomposition: Scenes detected from frame embeddings (collection source) enabled: true feature_extractor: feature_extractor_name: scene_extractor field_passthrough: - source_path: campaign_id input_mappings: embedding: multimodal_extractor_v1_embedding version: v1 input_schema: properties: campaign_id: type: string multimodal_extractor_v1_embedding: type: array output_schema: properties: campaign_id: type: string scene_extractor_v1_embedding: type: array source: collection_id: col_video_frames type: collection PaginationResponse: properties: total: anyOf: - type: integer - type: 'null' title: Total page: anyOf: - type: integer - type: 'null' title: Page page_size: anyOf: - type: integer - type: 'null' title: Page Size total_pages: anyOf: - type: integer - type: 'null' title: Total Pages next_page: anyOf: - type: string - type: 'null' title: Next Page previous_page: anyOf: - type: string - type: 'null' title: Previous Page next_cursor: anyOf: - type: string - type: 'null' title: Next Cursor type: object title: PaginationResponse description: |- PaginationResponse. Cursor-based pagination response: - Use next_cursor for navigation - Total count fields only populated when include_total=true CollectionListStats: properties: total_documents: type: integer title: Total Documents description: Total number of documents across all collections default: 0 avg_documents_per_collection: type: number title: Avg Documents Per Collection description: Average number of documents per collection default: 0 collections_with_taxonomies: type: integer title: Collections With Taxonomies description: Number of collections with taxonomy applications default: 0 total_feature_extractors: type: integer title: Total Feature Extractors description: Total number of feature extractors across all collections default: 0 total_taxonomies: type: integer title: Total Taxonomies description: Total number of taxonomy connections across all collections default: 0 total_retrievers: type: integer title: Total Retrievers description: Total number of retriever connections across all collections default: 0 type: object title: CollectionListStats description: Aggregate statistics for a list of collections. ErrorDetail: properties: message: type: string title: Message description: Human-readable error message type: type: string title: Type description: Stable error type identifier (machine-readable) code: anyOf: - type: string - type: 'null' title: Code description: >- Fine-grained error code for programmatic handling (e.g., namespace_name_taken, feature_extractor_not_found). Present only when consumers may need to branch on a specific error condition. details: anyOf: - additionalProperties: true type: object - type: 'null' title: Details description: >- Optional structured details to help debugging (validation errors, IDs, etc.) type: object required: - message - type title: ErrorDetail description: Error detail model. ValidationError: properties: loc: items: anyOf: - type: string - type: integer type: array title: Location msg: type: string title: Message type: type: string title: Error Type type: object required: - loc - msg - type title: ValidationError SortDirection: type: string enum: - asc - desc title: SortDirection description: Sort direction options. BucketSchema-Output: properties: properties: additionalProperties: $ref: '#/components/schemas/BucketSchemaField-Output' type: object title: Properties description: >- REQUIRED. Map of field names to their type definitions. Each field must have a 'type' from the supported types: metadata types (string, number, integer, float, boolean, object, array, date, datetime) or file/blob types (text, image, audio, video, pdf, excel). NOTE: Use Mixpeek types, NOT JSON Schema types — e.g. use 'string' not 'keyword', 'text' for text blobs, 'image' for image blobs. Example: {"title": {"type": "string"}, "photo": {"type": "image"}} additionalProperties: true type: object required: - properties title: BucketSchema description: >- Schema definition for bucket objects. IMPORTANT: The bucket schema defines what fields your bucket objects will have. This schema is REQUIRED if you want to: 1. Create collections that use input_mappings to process your bucket data 2. Validate object structure before ingestion 3. Enable type-safe data pipelines The schema defines the custom fields that will be used in: - Blob properties (e.g., "content", "thumbnail", "transcript") - Object metadata structure - Blob data structures Example workflow: 1. Create bucket WITH schema defining your data structure 2. Upload objects that conform to that schema 3. Create collections that map schema fields to feature extractors Without a bucket_schema, collections cannot use input_mappings. shared__collection__features__extractors__models__FeatureExtractorConfig-Output: properties: feature_extractor_name: type: string title: Feature Extractor Name description: Name of the feature extractor version: type: string title: Version description: Version of the feature extractor (e.g., 'v1', 'v2') params: anyOf: - additionalProperties: true type: object - type: 'null' title: Params description: >- Optional extractor parameters that affect vector index configuration. Parameters set here are locked at namespace creation and determine vector dimensions in Qdrant. Collections using this extractor must use compatible params. Example: {'model': 'siglip_base'} parameters: anyOf: - $ref: '#/components/schemas/AudioFingerprintExtractorParams' - $ref: '#/components/schemas/DocumentGraphExtractorParams' - $ref: '#/components/schemas/FaceIdentityExtractorParams' - $ref: '#/components/schemas/GeminiMultifileExtractorParams' - $ref: '#/components/schemas/ImageExtractorParams' - $ref: '#/components/schemas/MultimodalExtractorParams' - $ref: '#/components/schemas/PassthroughExtractorParams' - $ref: '#/components/schemas/ScrollingTextExtractorParams' - $ref: '#/components/schemas/TextExtractorParams' - $ref: '#/components/schemas/UniversalExtractorParams' - $ref: '#/components/schemas/WebScraperExtractorParams' - $ref: '#/components/schemas/CustomPluginParams' - type: 'null' title: Parameters description: >- Parameters for the feature extractor. Each extractor type has specific parameters. See the schema for your chosen extractor (e.g., MultimodalExtractorParams for multimodal_extractor). examples: - split_method: time time_split_interval: 10 - chunk_size: 512 split_by: sentences input_mappings: anyOf: - additionalProperties: anyOf: - type: string - items: type: string type: array type: object - items: $ref: '#/components/schemas/InputMapping' type: array title: Input Mappings description: >- Mapping from extractor input names to source field paths. Tells the extractor which source fields to process. Single value: {'image': 'thumbnail_url'} maps one blob field to one extractor input. Array value: {'files': ['image', 'spec_pdf', 'description']} maps multiple blob fields to a single extractor input as a list — used by multi-file extractors like gemini_multifile_extractor that embed all blobs of an object into one embedding. examples: - image: product_image text: title - video: video_url - text: content - files: - hero_image - spec_sheet - description field_passthrough: items: $ref: '#/components/schemas/FieldPassthrough' type: array title: Field Passthrough description: >- NOT REQUIRED. List of specific fields to pass through from source to output documents. These fields are included alongside extractor-computed features (embeddings, detections, etc.). Empty list = only extractor outputs in documents (default behavior). With entries = specified fields + extractor outputs in documents. How It Works: 1. During processing, fields are extracted from source object/document 2. They appear in output documents at the root level 3. Field filtering happens automatically (only listed fields included) 4. Use target_path to rename fields for cleaner schemas Common Use Cases: - Preserve identifiers: campaign_id, product_sku, order_id - Keep metadata: category, tags, author, created_at - Enable filtering: department, status, priority, region - Maintain context: title, description, source_url Behavior: - Works with include_all_source_fields=False (default): ONLY these fields included - Works with include_all_source_fields=True: These configs used for renaming/defaults - Fields must exist in source bucket_schema or upstream collection output_schema - Missing optional fields are omitted (unless default provided) - Missing required fields cause processing errors Output Schema: output_schema = field_passthrough fields + extractor output fields Example: ['title', 'category', 'text_extractor_v1_embedding'] examples: - - required: true source_path: title - required: true source_path: campaign_id - - default: uncategorized source_path: category - source_path: metadata.author target_path: author include_all_source_fields: type: boolean title: Include All Source Fields description: >- NOT REQUIRED. Whether to include ALL fields from source object/document in output. Default: False (only field_passthrough fields included). When False (RECOMMENDED): - Only fields listed in field_passthrough are included in output - Creates clean, predictable output schemas - Prevents data leakage of unwanted fields - Output = field_passthrough fields + extractor outputs When True (USE WITH CAUTION): - ALL source fields are included in output documents - field_passthrough still used for renaming/defaults/requirements - Can result in large documents if source has many fields - Can leak sensitive or unnecessary data - Output = all source fields + extractor outputs Use True When: - You want to preserve complete source data - Source has limited, well-defined fields - Downstream processing needs all context Use False When (MOST CASES): - You want clean, controlled output schemas - Source has many fields you don't need - You want explicit field selection - You're concerned about document size Examples: False: source={a,b,c,d} + passthrough=[a,b] → output={a,b,embedding} True: source={a,b,c,d} + passthrough=[a→x] → output={x,b,c,d,embedding} default: false feature_extractor_id: type: string title: Feature Extractor Id description: >- Construct unique identifier for the feature extractor instance (name + version). readOnly: true additionalProperties: false type: object required: - feature_extractor_name - version - feature_extractor_id title: FeatureExtractorConfig description: >- Configuration for a feature extractor with field passthrough support. A feature extractor processes source data (from buckets or collections) and produces features (embeddings, extracted text, detected objects, etc.). With field passthrough, you can also include selected source fields in the output documents alongside the computed features. Core Concepts: 1. **Feature Extraction**: Extractors compute features from input data (e.g., text → embeddings, image → detections, video → scenes) 2. **Field Passthrough**: Selectively preserve source fields in output (e.g., title, category, campaign_id from source → output documents) 3. **Output Schema**: Combination of passed-through fields + extractor outputs (e.g., {title, category, text_embedding} all in one document) How Field Passthrough Works: 1. Define which source fields to include via field_passthrough list 2. During processing, these fields are extracted from source 3. They appear in output documents at root level 4. Combine with extractor outputs for complete documents 5. Use target_path to rename fields for cleaner schemas Field Selection Modes: - **Explicit** (field_passthrough + include_all=False): Only listed fields pass through. Clean, controlled output. Example: passthrough=[title, category] → output has ONLY title, category, embedding - **Inclusive** (include_all=True): All source fields pass through, field_passthrough for renaming. Example: source has 10 fields → output has all 10 + embedding - **None** (no field_passthrough): Only extractor outputs in documents. Example: → output has ONLY embedding (no source fields) Use Cases: - **Preserve Identifiers**: Keep campaign_id, product_sku, order_id for tracking - **Enable Filtering**: Pass category, status, department for query filters - **Maintain Context**: Include title, description for display - **Track Metadata**: Preserve author, created_at, source for lineage - **Business Logic**: Keep priority, region, type for application logic Common Patterns: 1. **Minimal Passthrough** (recommended): field_passthrough=[{"source_path": "id"}], include_all=False → Clean output, only ID + extractor features 2. **Metadata Preservation**: field_passthrough=[ {"source_path": "title"}, {"source_path": "category"}, {"source_path": "created_at"} ] → Document has context for display and filtering 3. **Field Renaming**: field_passthrough=[ {"source_path": "doc_title", "target_path": "title"}, {"source_path": "metadata.author", "target_path": "author"} ] → Cleaner output schema with flattened fields 4. **Required Fields**: field_passthrough=[ {"source_path": "campaign_id", "required": True}, {"source_path": "priority", "default": 0} ] → Ensures critical fields always present Requirements: - feature_extractor_name: REQUIRED - name of the extractor - version: REQUIRED - extractor version (e.g., "v1") - parameters: NOT REQUIRED - extractor-specific config (model, thresholds, etc.) - input_mappings: NOT REQUIRED - maps extractor inputs to source fields - field_passthrough: NOT REQUIRED - which source fields to preserve (default: none) - include_all_source_fields: NOT REQUIRED - preserve all fields (default: false) examples: - description: Text extractor with field passthrough feature_extractor_name: text_extractor field_passthrough: - required: true source_path: title - source_path: author input_mappings: text: content parameters: model: text-embedding-3-small version: v1 - description: Video extractor with campaign metadata feature_extractor_name: multimodal_extractor field_passthrough: - source_path: campaign_id - source_path: duration_seconds input_mappings: video: video_url parameters: fps: 1 version: v1 - description: Image extractor with all source fields feature_extractor_name: image_extractor include_all_source_fields: true input_mappings: image: product_image parameters: model: clip-vit-base-patch32 version: v1 - description: Text extractor with field renaming feature_extractor_name: text_extractor field_passthrough: - source_path: metadata.author target_path: author - source_path: metadata.created_at target_path: created_at input_mappings: text: content version: v1 SourceConfig-Output: properties: type: $ref: '#/components/schemas/SourceType' description: >- REQUIRED. Type of source for this collection. 'bucket': Process objects from one or more buckets (first-stage processing). 'collection': Process documents from another collection (downstream processing). Use 'bucket' for initial data ingestion, 'collection' for decomposition trees. examples: - bucket - collection bucket_ids: anyOf: - items: type: string type: array minItems: 1 - type: 'null' title: Bucket Ids description: >- List of bucket IDs when type='bucket'. REQUIRED when type='bucket'. NOT ALLOWED when type='collection'. Can specify one or more buckets to process. Single bucket: Use array with one element ['bkt_id']. Multiple buckets: All buckets MUST have compatible schemas. Schema compatibility validated at collection creation. Compatible schemas have: 1) Same field names, 2) Same field types, 3) Same required status. Documents will include root_bucket_id to track which bucket they came from. Use cases: multi-region data, multi-team consolidation, environment aggregation. examples: - - bkt_marketing_videos - - bkt_us_products - bkt_eu_products - bkt_asia_products - - bkt_staging_data - bkt_production_data source_namespace_id: anyOf: - type: string - type: 'null' title: Source Namespace Id description: >- Namespace ID where the source buckets reside. Use this to process buckets from a different namespace within the same organization. When omitted, buckets are looked up in the current (collection's) namespace. Only valid when type='bucket'. collection_id: anyOf: - type: string - type: 'null' title: Collection Id description: >- Collection ID when type='collection' (single collection). Use this OR collection_ids (not both). REQUIRED when type='collection' and processing single collection. NOT ALLOWED when type='bucket'. The collection will process documents from this upstream collection. The upstream collection's output_schema becomes this collection's input_schema. This enables decomposition trees (multi-stage pipelines). Example: Process frames collection → create scenes collection. examples: - col_video_frames - col_book_chapters - col_product_images collection_ids: anyOf: - items: type: string type: array minItems: 1 - type: 'null' title: Collection Ids description: >- List of collection IDs when type='collection' (multiple collections). Use this OR collection_id (not both). REQUIRED when type='collection' and processing multiple collections. NOT ALLOWED when type='bucket'. Used for operations that consolidate multiple upstream collections. Example: Clustering across multiple collections → cluster output collection. All collections must have compatible schemas for consolidation operations. examples: - - col_us_products - col_eu_products - - col_images_2023 - col_images_2024 inherited_bucket_ids: anyOf: - items: type: string type: array - type: 'null' title: Inherited Bucket Ids description: >- List of original bucket IDs that source collections originated from. OPTIONAL. Only used when type='collection'. Tracks the complete lineage chain: buckets → collections → derived collections. Extracted from upstream collection metadata at collection creation time. Enables tracing derived collections (like cluster outputs) back to original data sources. Example: Cluster output collection inherits bucket IDs from its source collections. Format: List of bucket IDs with 'bkt_' prefix. examples: - - bkt_marketing_videos - - bkt_us_products - bkt_eu_products - null source_filters: anyOf: - $ref: '#/components/schemas/SourceFilters-Output' - type: 'null' description: >- Optional filters to apply to source data. When specified, only objects/documents matching these filters will be processed by this collection. Filters are evaluated at batch creation time. Uses same LogicalOperator model as list APIs for consistency. examples: - filters: AND: - field: blobs.type operator: eq value: video - filters: AND: - field: metadata.status operator: eq value: active - {} type: object required: - type title: SourceConfig description: >- Configuration for collection source (bucket(s) or collection). Collections can process data from two types of sources: 1. **Bucket Source**: Process raw objects from one or more buckets (first-stage processing) - Use this to create your initial collections from uploaded data - Can specify multiple buckets to consolidate data from different sources - All buckets must have compatible schemas (validated at creation) - Example: Videos from multiple regions → Frame extraction collection 2. **Collection Source**: Process documents from another collection (decomposition trees) - Use this to create multi-stage processing pipelines - Example: Frames collection → Scene detection collection Multi-Bucket Requirements: - All buckets must have compatible schemas (same fields, types, and required status) - Schema compatibility is validated when the collection is created - Documents track which specific bucket they came from via root_bucket_id - Useful for consolidating data from multiple regions, teams, or environments The source determines: - What data the feature extractor receives as input - The input_schema available for input_mappings and field_passthrough - The lineage tracking in output documents Examples: Single bucket: {"type": "bucket", "bucket_ids": ["bkt_products"]} Multi-bucket: {"type": "bucket", "bucket_ids": ["bkt_us", "bkt_eu", "bkt_asia"]} Collection: {"type": "collection", "collection_id": "col_frames"} examples: - bucket_ids: - bkt_marketing_videos description: Single bucket source type: bucket - bucket_ids: - bkt_marketing_ads description: Bucket source with video-only filter source_filters: filters: AND: - field: blobs.type operator: eq value: video type: bucket - bucket_ids: - bkt_us_products - bkt_eu_products - bkt_asia_products description: Multi-bucket source (region consolidation) type: bucket - bucket_ids: - bkt_us_ads - bkt_eu_ads - bkt_asia_ads description: Multi-bucket source with brand filter source_filters: filters: OR: - field: brand_name operator: in value: - Acme - TechCo - field: category operator: eq value: premium type: bucket - collection_id: col_video_frames description: Collection source (decomposition tree) type: collection - collection_id: col_raw_frames description: Collection source with active documents only source_filters: filters: AND: - field: __fully_enriched operator: eq value: true - field: quality_score operator: gte value: 0.8 type: collection - collection_ids: - col_us_products - col_eu_products - col_asia_products description: Multi-collection source (cluster output) type: collection SingleLineageEntry: properties: source_config: $ref: '#/components/schemas/SourceConfig-Output' description: Configuration of the source for this lineage entry feature_extractor: $ref: >- #/components/schemas/shared__collection__features__extractors__models__FeatureExtractorConfig-Output description: Single feature extractor applied at this stage output_schema: $ref: '#/components/schemas/BucketSchema-Output' description: Output schema produced by this processing stage type: object required: - source_config - feature_extractor - output_schema title: SingleLineageEntry description: >- Single entry in the lineage chain of a collection. Each lineage entry represents one processing stage with one feature extractor. CollectionScheduleConfig: properties: trigger_type: type: string title: Trigger Type description: 'Schedule type: ''cron'' or ''interval''' examples: - cron - interval schedule_config: additionalProperties: true type: object title: Schedule Config description: >- Schedule configuration. For cron: {cron_expression, timezone}. For interval: {interval_seconds, start_immediately}. examples: - cron_expression: 0 2 * * * timezone: UTC - interval_seconds: 21600 start_immediately: false description: anyOf: - type: string - type: 'null' title: Description description: Human-readable description of the schedule. type: object required: - trigger_type - schedule_config title: CollectionScheduleConfig description: |- Schedule configuration for automatic collection re-processing. Attaches a cron or interval schedule to a collection, which creates a COLLECTION_TRIGGER trigger behind the scenes. This is a DX convenience so users don't need to create triggers manually. Examples: Daily re-crawl at 2am UTC: {"trigger_type": "cron", "schedule_config": {"cron_expression": "0 2 * * *"}} Every 6 hours: {"trigger_type": "interval", "schedule_config": {"interval_seconds": 21600}} TieringRule: properties: rule_type: type: string enum: - auto_evict - auto_archive - auto_rehydrate title: Rule Type enabled: type: boolean title: Enabled default: false threshold_days: anyOf: - type: integer - type: 'null' title: Threshold Days type: object required: - rule_type title: TieringRule description: 'A single automatic storage tiering rule (V1: stored but not enforced).' TaxonomyApplicationConfig-Output: properties: taxonomy_id: type: string title: Taxonomy Id description: ID of the `TaxonomyModel` to execute. execution_mode: $ref: '#/components/schemas/TaxonomyExecutionMode' description: >- Execution mode for taxonomy enrichment. Materializes results during ingestion. default: materialize target_collection_id: anyOf: - type: string - type: 'null' title: Target Collection Id description: >- Optional collection to persist results into when `execution_mode` is 'materialize'. If omitted, the source collection is updated in-place. scroll_filters: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - type: 'null' description: >- Additional filters applied when scrolling the source collection before enrichment. execution_phase: type: integer maximum: 3 minimum: 1 title: Execution Phase description: >- Which phase this taxonomy runs in. Default: 1 (TAXONOMY phase, runs first). Valid values: 1=TAXONOMY, 2=CLUSTER, 3=ALERT. Lower phases run earlier. default: 1 priority: type: integer maximum: 1000 minimum: 0 title: Priority description: Priority within the execution phase (higher = runs first) default: 0 hierarchical_enrichment_style: $ref: '#/components/schemas/HierarchicalEnrichmentStyle' description: >- For hierarchical taxonomies, controls how enrichment fields are structured. 'full_chain': Each tier writes to tier-prefixed fields (tier1_id, tier2_id, etc.). 'best_match': Only deepest match stored (category_id, category_name, path[]). 'combined' (default): Both full chain AND best match summary fields. default: combined additionalProperties: false type: object required: - taxonomy_id title: TaxonomyApplicationConfig description: |- Configuration block that attaches a taxonomy to a collection. Supports execution phase ordering for unified post-processing with taxonomies, clusters, and alerts. examples: - execution_mode: materialize taxonomy_id: tax_abc123 - execution_mode: materialize execution_phase: 1 priority: 10 target_collection_id: col_enriched_v1 taxonomy_id: tax_abc123 - execution_mode: materialize hierarchical_enrichment_style: full_chain taxonomy_id: tax_iab_content ClusterApplicationConfig: properties: cluster_id: type: string title: Cluster Id description: >- ID of the cluster to execute (must exist and use this collection as input) auto_execute_on_batch: type: boolean title: Auto Execute On Batch description: >- Automatically execute cluster when batch processing completes for this collection. If False, cluster must be executed manually via API. default: true min_document_threshold: anyOf: - type: integer - type: 'null' title: Min Document Threshold description: >- Minimum number of documents required before executing cluster. If document_count < threshold, clustering is skipped. Useful to avoid clustering on small datasets. cooldown_seconds: type: integer title: Cooldown Seconds description: >- Minimum time (in seconds) between automatic cluster executions. Prevents excessive re-clustering on frequent batch completions. Default: 3600 seconds (1 hour). default: 3600 execution_phase: type: integer maximum: 3 minimum: 1 title: Execution Phase description: >- Which phase this cluster runs in. Default: 2 (CLUSTER phase, after taxonomies). Valid values: 1=TAXONOMY, 2=CLUSTER, 3=ALERT. Lower phases run earlier. default: 2 priority: type: integer maximum: 1000 minimum: 0 title: Priority description: Priority within the execution phase (higher = runs first) default: 0 type: object required: - cluster_id title: ClusterApplicationConfig description: >- Configuration for automatic cluster execution on collection. Similar to TaxonomyApplicationConfig, this attaches a cluster to a collection and defines when/how it should be automatically executed. Used in CollectionModel.cluster_applications field. Supports execution phase ordering for unified post-processing with taxonomies, clusters, and alerts. examples: - auto_execute_on_batch: true cluster_id: clust_product_categories cooldown_seconds: 3600 min_document_threshold: 100 - auto_execute_on_batch: true cluster_id: clust_user_segments cooldown_seconds: 7200 execution_phase: 2 min_document_threshold: 500 priority: 10 AlertApplicationConfig-Output: properties: alert_id: type: string title: Alert Id description: ID of the alert to execute examples: - alt_safety_001 - alt_content_moderation execution_mode: $ref: '#/components/schemas/AlertExecutionMode' description: When this alert should execute default: on_ingest input_mappings: items: $ref: '#/components/schemas/AlertInputMapping' type: array minItems: 1 title: Input Mappings description: Map document fields or constants to retriever input parameters execution_phase: $ref: '#/components/schemas/PostProcessingPhase' description: >- Which phase this alert runs in. Default: ALERT (phase 3, after taxonomies and clusters) default: 3 priority: type: integer maximum: 1000 minimum: 0 title: Priority description: Priority within the execution phase (higher = runs first) default: 0 additionalProperties: false type: object required: - alert_id - input_mappings title: AlertApplicationConfig description: |- Configuration for attaching an alert to a collection. The key responsibility here is INPUT MAPPING: connecting document fields (or constants) to the retriever's expected inputs. Note: Filtering logic (scroll_filters, min_score, etc.) belongs in the retriever, not here. The retriever owns all query semantics. Use Cases: - Attach safety alert to video upload collection - Configure different field mappings for different collections - Set execution priority for multiple alerts Attributes: alert_id: ID of the alert to execute execution_mode: When this alert should execute input_mappings: Map document fields or constants to retriever inputs execution_phase: Which phase this alert runs in (default: ALERT) priority: Priority within the execution phase (higher = runs first) examples: - alert_id: alt_safety_001 execution_mode: on_ingest execution_phase: 3 input_mappings: - input_key: query_embedding source: path: features.video_embedding source_type: document_field - input_key: target_collection source: source_type: constant value: col_known_incidents priority: 100 RetrieverEnrichmentConfig-Output: properties: retriever_id: type: string title: Retriever Id description: ID of the retriever to execute examples: - ret_classifier_001 - ret_cross_reference input_mappings: items: $ref: '#/components/schemas/EnrichmentInputMapping' type: array minItems: 1 title: Input Mappings description: Map document fields or constants to retriever input parameters write_back_fields: items: $ref: '#/components/schemas/WriteBackFieldMapping' type: array minItems: 1 title: Write Back Fields description: Which retriever result fields to write back to the document execution_phase: $ref: '#/components/schemas/PostProcessingPhase' description: >- Which phase this enrichment runs in. Default: RETRIEVER_ENRICHMENT (phase 4, after taxonomies, clusters, and alerts) default: 4 priority: type: integer maximum: 1000 minimum: 0 title: Priority description: Priority within the execution phase (higher = runs first) default: 0 scroll_filters: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - type: 'null' description: Optional filters to select which documents to enrich enabled: type: boolean title: Enabled description: Whether this enrichment is active default: true type: object required: - retriever_id - input_mappings - write_back_fields title: RetrieverEnrichmentConfig description: |- Configuration for attaching a retriever enrichment to a collection. Retriever enrichments run a retriever pipeline on each document during post-processing and write selected result fields back to the document. Attributes: retriever_id: ID of the retriever to execute input_mappings: How to map document fields to retriever inputs write_back_fields: Which result fields to write back to the document execution_phase: Which post-processing phase to run in (default: RETRIEVER_ENRICHMENT) priority: Priority within the execution phase (higher = runs first) scroll_filters: Optional filters to select which documents to enrich enabled: Whether this enrichment is active examples: - enabled: true input_mappings: - input_key: query source: path: title source_type: document_field retriever_id: ret_classifier_001 write_back_fields: - mode: first source_field: category target_field: _enrichment_category BucketSchemaField-Output: properties: type: $ref: '#/components/schemas/BucketSchemaFieldType' default: anyOf: - {} - type: 'null' title: Default items: anyOf: - $ref: '#/components/schemas/BucketSchemaField-Output' - type: 'null' properties: anyOf: - additionalProperties: $ref: '#/components/schemas/BucketSchemaField-Output' type: object - type: 'null' title: Properties examples: anyOf: - items: {} type: array - type: 'null' title: Examples description: >- OPTIONAL. List of example values for this field. Used by Apps to show example inputs in the UI. Provide multiple diverse examples when possible. description: anyOf: - type: string - type: 'null' title: Description enum: anyOf: - items: {} type: array - type: 'null' title: Enum required: anyOf: - type: boolean - type: 'null' title: Required default: false additionalProperties: true type: object required: - type title: BucketSchemaField description: Schema field definition for bucket objects. AudioFingerprintExtractorParams: properties: extractor_type: type: string const: audio_fingerprint_extractor title: Extractor Type description: Discriminator field. Must be 'audio_fingerprint_extractor'. default: audio_fingerprint_extractor segment_duration_sec: type: number maximum: 30 minimum: 1 title: Segment Duration Sec description: >- Duration of each audio segment in seconds. 5.0: Recommended for sound mark matching. Shorter segments increase recall but reduce per-segment context. default: 5 segment_hop_sec: type: number maximum: 15 minimum: 0.5 title: Segment Hop Sec description: >- Hop size between segments in seconds. 2.5: 50% overlap (recommended). Set equal to segment_duration_sec for no overlap. default: 2.5 sample_rate: type: integer title: Sample Rate description: >- Target sample rate for audio. 48000: CLAP default (recommended). Audio is resampled to this rate before embedding. default: 48000 normalize_embeddings: type: boolean title: Normalize Embeddings description: >- L2-normalize embeddings to unit vectors (recommended for cosine similarity). default: true max_audio_length_sec: type: number maximum: 600 minimum: 1 title: Max Audio Length Sec description: >- Maximum audio length to process in seconds. 120: Default (2 minutes). Audio beyond this is truncated. default: 120 type: object title: AudioFingerprintExtractorParams description: |- Parameters for the Audio Fingerprint Extractor. Processes audio files (or audio extracted from video) through CLAP (Contrastive Language-Audio Pretraining) to produce 512-d embeddings suitable for audio fingerprint matching. Core Pipeline: 1. Audio extraction (if video input, via FFmpeg) 2. Segmentation into fixed-length windows 3. CLAP embedding (laion/clap-htsat-tiny, 512-d) 4. L2 normalization Use Cases: - Sound mark detection (IP safety) - Audio similarity search - Music/jingle identification - Audio deduplication examples: - description: Sound mark detection (IP safety) extractor_type: audio_fingerprint_extractor normalize_embeddings: true sample_rate: 48000 segment_duration_sec: 5 segment_hop_sec: 2.5 - description: Music identification (longer segments) extractor_type: audio_fingerprint_extractor sample_rate: 48000 segment_duration_sec: 10 segment_hop_sec: 5 DocumentGraphExtractorParams: properties: extractor_type: type: string const: document_graph_extractor title: Extractor Type description: >- Discriminator field for parameter type identification. Must be 'document_graph_extractor'. default: document_graph_extractor use_layout_detection: type: boolean title: Use Layout Detection description: >- Enable ML-based layout detection to find ALL document elements (text, images, tables, figures). When enabled, uses the configured layout_detector to detect and extract both text regions AND non-text elements (scanned images, figures, charts) as separate documents. **Recommended for**: Scanned documents, image-heavy PDFs, mixed content documents. **When disabled**: Falls back to text-only extraction (faster but misses images). Default: True (detects all elements including images). default: true layout_detector: type: string enum: - pymupdf - docling title: Layout Detector description: >- Layout detection engine to use when use_layout_detection=True. 'pymupdf': Fast, rule-based detection using PyMuPDF heuristics (~15 pages/sec). 'docling': SOTA ML-based detection using IBM Docling with DiT model (~3-8 sec/doc). **Docling advantages**: Better semantic type detection (section_header vs paragraph), true table structure extraction (rows/cols), more accurate figure detection. **PyMuPDF advantages**: Much faster, lower memory usage, simpler dependencies. Default: 'pymupdf' for speed. Use 'docling' for accuracy-critical applications. default: pymupdf vertical_threshold: type: number maximum: 100 minimum: 1 title: Vertical Threshold description: >- Maximum vertical gap (in points) between lines to be grouped in same block. Increase for looser grouping, decrease for tighter blocks. Default 15pt works well for standard documents. default: 15 horizontal_threshold: type: number maximum: 200 minimum: 1 title: Horizontal Threshold description: >- Maximum horizontal distance (in points) for overlap detection. Affects column detection and block merging. Increase for wider columns, decrease for narrow layouts. default: 50 min_text_length: type: integer maximum: 500 minimum: 1 title: Min Text Length description: >- Minimum text length (characters) to keep a block. Blocks with less text are filtered out. Helps remove noise and tiny fragments. default: 20 base_confidence: type: number maximum: 1 minimum: 0 title: Base Confidence description: >- Base confidence score for embedded (native) text. Penalties are subtracted for OCR artifacts, encoding issues, etc. default: 0.85 min_confidence_for_vlm: type: number maximum: 1 minimum: 0 title: Min Confidence For Vlm description: >- Confidence threshold below which VLM correction is triggered. Blocks with confidence < this value get sent to VLM for correction. Only applies when use_vlm_correction=True. default: 0.6 use_vlm_correction: type: boolean title: Use Vlm Correction description: >- Enable VLM (Vision Language Model) correction for low-confidence blocks. Uses Gemini/GPT-4V to correct OCR errors by analyzing the page image. Significantly slower (~1 page/sec) but improves accuracy for degraded docs. default: true fast_mode: type: boolean title: Fast Mode description: >- Skip VLM correction entirely for maximum throughput (~15 pages/sec). Overrides use_vlm_correction. Use when speed is more important than accuracy. default: false vlm_provider: type: string title: Vlm Provider description: >- LLM provider for VLM correction. Options: 'google' (Gemini), 'openai' (GPT-4V), 'anthropic' (Claude). Google recommended for best vision quality. default: google vlm_model: type: string title: Vlm Model description: >- Specific model for VLM correction. Examples: 'gemini-2.5-flash', 'gpt-4o', 'claude-3-5-sonnet'. default: gemini-2.5-flash llm_api_key: anyOf: - type: string - type: 'null' title: Llm Api Key description: >- API key for VLM correction (BYOK - Bring Your Own Key). Supports: - Direct key: 'sk-proj-abc123...' - Secret reference: '{{SECRET.openai_api_key}}' When using secret reference, the key is loaded from your organization's secrets vault at runtime. Store secrets via POST /v1/organizations/secrets. If not provided, uses Mixpeek's default API keys. run_text_embedding: type: boolean title: Run Text Embedding description: >- Generate text embeddings for semantic search over block content. Uses E5-Large (1024-dim) for multilingual support. default: true render_dpi: type: integer maximum: 300 minimum: 72 title: Render Dpi description: >- DPI for page rendering (used for VLM correction). 72: Fast, lower quality. 150: Balanced (recommended). 300: High quality, slower. default: 150 generate_thumbnails: type: boolean title: Generate Thumbnails description: >- Generate thumbnail images for blocks. Useful for visual previews and UI display. default: true thumbnail_mode: type: string title: Thumbnail Mode description: >- Thumbnail generation mode. 'full_page': Low-res thumbnail of entire page. 'segment': Cropped thumbnail of just the block's bounding box. 'both': Generate both types (recommended for flexibility). default: both thumbnail_dpi: type: integer maximum: 150 minimum: 36 title: Thumbnail Dpi description: >- DPI for thumbnail generation. Lower DPI = smaller files. 72: Standard web quality. 36: Very small thumbnails. default: 72 type: object title: DocumentGraphExtractorParams description: >- Parameters for the document graph extractor. This extractor decomposes PDFs into spatial blocks with layout classification, confidence scoring, and optional VLM correction for degraded documents. **When to Use**: - Historical/archival document processing (FBI files, old records) - Scanned documents with mixed quality - Documents requiring spatial understanding (forms, tables, multi-column) - When you need block-level granularity with bounding boxes - When confidence scoring is needed for downstream filtering **When NOT to Use**: - Simple text-only documents -> Use text_extractor instead - When page-level granularity is sufficient -> Use pdf_extractor instead - Real-time processing requirements -> VLM correction adds latency examples: - description: Fast processing mode (no VLM, maximum throughput) extractor_type: document_graph_extractor fast_mode: true generate_thumbnails: true layout_detector: pymupdf run_text_embedding: true use_case: >- High-volume document ingestion where speed matters more than perfect accuracy use_layout_detection: true - description: Archival documents with VLM correction (recommended for old scans) extractor_type: document_graph_extractor layout_detector: pymupdf min_confidence_for_vlm: 0.6 render_dpi: 150 run_text_embedding: true use_case: >- Historical archives, FBI files, old scanned documents with degraded quality use_layout_detection: true use_vlm_correction: true vlm_model: gemini-2.5-flash vlm_provider: google - description: SOTA accuracy mode with Docling (best for tables/figures) extractor_type: document_graph_extractor fast_mode: true generate_thumbnails: true layout_detector: docling run_text_embedding: true use_case: >- Documents with complex tables, figures, or requiring accurate semantic typing use_layout_detection: true FaceIdentityExtractorParams: properties: extractor_type: type: string const: face_identity_extractor title: Extractor Type description: >- Discriminator field for parameter type identification. Must be 'face_identity_extractor'. default: face_identity_extractor detection_model: type: string enum: - scrfd_500m - scrfd_2.5g - scrfd_10g title: Detection Model description: >- SCRFD model for face detection. 'scrfd_500m': Fastest (2-3ms). 'scrfd_2.5g': Balanced (5-7ms), recommended. 'scrfd_10g': Highest accuracy (10-15ms). default: scrfd_2.5g min_face_size: type: integer maximum: 200 minimum: 10 title: Min Face Size description: >- Minimum face size in pixels to detect. 20px: Balanced. 40px: Higher quality. 10px: Maximum recall. default: 20 detection_threshold: type: number maximum: 1 minimum: 0 title: Detection Threshold description: Confidence threshold for face detection (0.0-1.0). default: 0.5 max_faces_per_image: anyOf: - type: integer minimum: 1 - type: 'null' title: Max Faces Per Image description: 'Maximum number of faces to process per image. None: Process all.' normalize_embeddings: type: boolean title: Normalize Embeddings description: L2-normalize embeddings to unit vectors (recommended). default: true enable_quality_scoring: type: boolean title: Enable Quality Scoring description: Compute quality scores (blur, size, landmarks). Adds ~5ms per face. default: true quality_threshold: anyOf: - type: number maximum: 1 minimum: 0 - type: 'null' title: Quality Threshold description: >- Minimum quality score to index faces. None: Index all faces. 0.5: Moderate filtering. 0.7: High quality only. max_video_length: type: integer maximum: 300 minimum: 1 title: Max Video Length description: >- Maximum video length in seconds. 60: Default. 10: Recommended for retrieval. 300: Maximum (extraction only). default: 60 video_sampling_fps: anyOf: - type: number maximum: 60 minimum: 0.1 - type: 'null' title: Video Sampling Fps description: >- Frames per second to sample from video. 1.0: One frame per second (recommended). default: 1 video_deduplication: type: boolean title: Video Deduplication description: >- Remove duplicate faces across video frames (extraction only). Reduces 90-95% redundancy. NOT used in retrieval. default: true video_deduplication_threshold: type: number maximum: 1 minimum: 0 title: Video Deduplication Threshold description: >- Cosine similarity threshold for deduplication. 0.8: Conservative (default). default: 0.8 output_mode: type: string enum: - per_face - per_image title: Output Mode description: >- 'per_face': One document per face (recommended). 'per_image': One doc per image with faces array. default: per_face include_face_crops: type: boolean title: Include Face Crops description: >- Include aligned 112×112 face crops as base64. Adds ~5KB per face. Required for LLM cluster labeling to see actual faces instead of hallucinating. default: true include_source_frame_thumbnail: type: boolean title: Include Source Frame Thumbnail description: >- Include resized source frame/image as base64 thumbnail (~15-30KB per face). Used for display with bounding box overlay. default: false store_detection_metadata: type: boolean title: Store Detection Metadata description: Store bbox, landmarks, detection scores. Recommended for debugging. default: true type: object title: FaceIdentityExtractorParams description: >- Parameters for the Face Identity Extractor. The Face Identity Extractor processes images or video frames to detect, align, and embed faces using production-grade SOTA models (SCRFD + ArcFace). Core Pipeline: 1. SCRFD Detection → Bounding boxes + 5 landmarks 2. 5-Point Affine Alignment → 112×112 canonical face 3. ArcFace Embedding → 512-d L2-normalized vector 4. Optional Quality Scoring → Filter low-quality faces Use Cases: - Face verification (1:1 matching) - Face identification (1:N search) - Face clustering (group photos by person) - Duplicate face detection examples: - description: Employee verification (high quality, 1:1 matching) detection_model: scrfd_2.5g detection_threshold: 0.7 enable_quality_scoring: true extractor_type: face_identity_extractor max_faces_per_image: 1 min_face_size: 40 normalize_embeddings: true output_mode: per_face quality_threshold: 0.5 use_case: Corporate access control, employee ID photos for badge matching - description: Photo library organization (multiple faces) detection_model: scrfd_2.5g detection_threshold: 0.6 enable_quality_scoring: true extractor_type: face_identity_extractor min_face_size: 30 output_mode: per_face store_detection_metadata: true use_case: 'Personal photo management: group photos by person' GeminiMultifileExtractorParams: properties: extractor_type: type: string const: gemini_multifile_extractor title: Extractor Type description: Discriminator field for parameter type identification. default: gemini_multifile_extractor output_dimensionality: type: integer maximum: 3072 minimum: 256 title: Output Dimensionality description: >- Output embedding dimensions. Gemini Embedding 2 supports 3072 (default), 768, or 256 via truncation. Lower dimensions reduce storage cost at slight quality loss. default: 3072 task_type: type: string title: Task Type description: >- Embedding intent used as a text instruction for Gemini Embedding 2. Common values: RETRIEVAL_DOCUMENT, RETRIEVAL_QUERY, SEMANTIC_SIMILARITY, CLASSIFICATION. default: RETRIEVAL_DOCUMENT input_key: type: string title: Input Key description: >- The input_mappings key whose value is the list of blob fields to embed together. Must match the key used in input_mappings (e.g., 'files'). Default: 'files'. default: files type: object title: GeminiMultifileExtractorParams description: >- Parameters for the Gemini Multifile Extractor. Uses Gemini Embedding 2 (gemini-embedding-2) to embed all files of an object into a single 3072-d vector in one API call. Supports images, video, audio, PDF, and text blobs. examples: - extractor_type: gemini_multifile_extractor input_key: files output_dimensionality: 3072 task_type: RETRIEVAL_DOCUMENT ImageExtractorParams: properties: extractor_type: type: string const: image_extractor title: Extractor Type description: Discriminator field for parameter type identification. default: image_extractor enable_thumbnails: type: boolean title: Enable Thumbnails description: Whether to generate thumbnail images. default: true use_cdn: type: boolean title: Use Cdn description: Whether to use CloudFront CDN for thumbnail delivery. default: false type: object title: ImageExtractorParams description: Parameters for the Image Extractor. examples: - enable_thumbnails: true extractor_type: image_extractor use_cdn: false MultimodalExtractorParams: properties: extractor_type: type: string const: multimodal_extractor title: Extractor Type description: Discriminator field. Must be 'multimodal_extractor'. default: multimodal_extractor split_method: $ref: '#/components/schemas/SplitMethod' description: Video splitting strategy. default: time description_prompt: type: string title: Description Prompt description: Prompt for description generation. default: >- Watch this video segment carefully and describe exactly what you see. Do not make up or infer details that are not visible in the footage. Include: who is shown (gender, appearance, actions), what they are doing, the setting/location, and any products, text, or branding visible on screen. time_split_interval: anyOf: - type: integer - type: 'null' title: Time Split Interval description: Interval in seconds for 'time' splitting. default: 10 silence_db_threshold: anyOf: - type: integer - type: 'null' title: Silence Db Threshold description: 'Decibel threshold for silence detection. Recommended: -40.' scene_detection_threshold: anyOf: - type: number - type: 'null' title: Scene Detection Threshold description: 'Scene detection threshold (0.0-1.0). Recommended: 0.5.' run_transcription: type: boolean title: Run Transcription description: Run Whisper transcription on segments. default: false transcription_language: type: string title: Transcription Language description: Transcription language code. default: en run_video_description: type: boolean title: Run Video Description description: Generate Gemini descriptions for segments. default: false run_transcription_embedding: type: boolean title: Run Transcription Embedding description: Generate E5 embeddings for transcriptions (1024D). default: false run_ocr_embedding: type: boolean title: Run Ocr Embedding description: Generate E5 embeddings for OCR text (1024D). Requires run_ocr. default: false run_description_embedding: type: boolean title: Run Description Embedding description: >- Generate E5 embeddings for descriptions (1024D). Requires run_video_description. default: false run_multimodal_embedding: type: boolean title: Run Multimodal Embedding description: >- Generate Gemini Embedding 2 multimodal embeddings (3072D). Creates unified embeddings across video, image, text, audio, and GIF content. default: true run_ocr: type: boolean title: Run Ocr description: Extract text from video frames via Gemini OCR. default: false max_segment_duration: anyOf: - type: number - type: 'null' title: Max Segment Duration description: >- Maximum duration in seconds for any single segment. Scene/silence segments longer than this are subdivided. Set to None to disable. Default: 30s. default: 30 sensitivity: type: string title: Sensitivity description: Scene detection sensitivity. default: low enable_thumbnails: type: boolean title: Enable Thumbnails description: Generate thumbnail images for segments. default: true use_cdn: type: boolean title: Use Cdn description: Use CloudFront CDN for thumbnail delivery. default: false generation_config: $ref: '#/components/schemas/GenerationConfig' output_dimensionality: type: integer title: Output Dimensionality description: >- Output embedding dimensions. Gemini Embedding 2 supports Matryoshka dimension reduction: 3072 (full), 1536, or 768. default: 3072 task_type: type: string title: Task Type description: >- Embedding task type hint. Options: RETRIEVAL_DOCUMENT, RETRIEVAL_QUERY, SEMANTIC_SIMILARITY, CLASSIFICATION. default: RETRIEVAL_DOCUMENT response_shape: anyOf: - type: string - additionalProperties: true type: object - type: 'null' title: Response Shape description: >- Custom structured output schema for Gemini extraction. String for natural language prompt, dict for explicit JSON schema. embedding_task: anyOf: - type: string - type: 'null' title: Embedding Task description: >- Embedding task hint for instruction-aware models (E5). Prefer setting this at collection level (embedding_task on the collection) rather than here. Collection-level overrides this value. Defaults to 'retrieval_document'. Values: retrieval_document, retrieval_query, semantic_similarity, classification, clustering. Note: Vertex AI multimodal embeddings ignore this — only E5 transcription embeddings use it. type: object title: MultimodalExtractorParams description: |- Parameters for multimodal extractor v2. Same pipeline as v1 but uses Gemini Embedding 2 (3072D) for the multimodal embedding step. Supports configurable output dimensions via Matryoshka representation learning (3072/1536/768). PassthroughExtractorParams: properties: extractor_type: type: string const: passthrough_extractor title: Extractor Type description: Discriminator field for parameter type identification. default: passthrough_extractor preserve_metadata: type: boolean title: Preserve Metadata description: Preserve source object metadata in output document. default: true type: object title: PassthroughExtractorParams description: |- Parameters for passthrough extractor. Minimal configuration - just passes data through with canonicalization. ScrollingTextExtractorParams: properties: extractor_type: type: string const: scrolling_text_extractor title: Extractor Type description: Discriminator field. Must be 'scrolling_text_extractor'. default: scrolling_text_extractor fps: type: number maximum: 30 minimum: 1 title: Fps description: >- Frame sampling rate for analysis. Higher values improve detection accuracy for fast-scrolling text but increase processing time. 5 FPS works well for most video ads and tickers. default: 5 strip_height: type: integer maximum: 200 minimum: 10 title: Strip Height description: >- Height (in pixels) of each scanning strip used for phase correlation. Should roughly match the height of the scrolling text band. Smaller values detect narrower text bands; larger values are more robust but may miss thin tickers. 40px works for most standard video ads. default: 40 min_shift_px: type: number maximum: 20 minimum: 0.5 title: Min Shift Px description: >- Minimum pixel shift per frame to consider a strip as 'scrolling'. Lower values detect slower-moving text; higher values filter out noise. 2.0px is a good default for 5 FPS sampling. default: 2 consistency_ratio: type: number maximum: 1 minimum: 0.3 title: Consistency Ratio description: >- Fraction of frame pairs that must show consistent shift for a band to be classified as scrolling. 0.6 means 60% of frames must agree. Lower values detect intermittent scrolling; higher values reduce false positives. default: 0.6 pad: type: integer maximum: 50 minimum: 0 title: Pad description: Pixel padding above/below detected band when cropping for stitching. default: 8 type: object title: ScrollingTextExtractorParams description: >- Parameters for the scrolling text extractor. Detects and extracts scrolling/marquee text from video using computer vision (phase-correlation band detection + panoramic stitching) and VLM-based OCR. **When to Use**: - Video ads with scrolling promotional banners or tickers - News broadcasts with scrolling chyrons or tickers - Videos with scrolling terms & conditions or disclaimers - Social media content with horizontally scrolling text overlays - Credits sequences in film/TV content - Live event streams with scrolling info bars - Any video where important text scrolls across the screen **When NOT to Use**: - Static text overlays → use multimodal_extractor with run_ocr=True - Spoken content → use multimodal_extractor with run_transcription=True - Text documents/PDFs → use text_extractor - Animated text that fades/scales (not linear scroll) → use multimodal_extractor **How It Works**: 1. Sample frames from video at configurable FPS 2. Split each frame into horizontal and vertical strips 3. Phase-correlate consecutive frames to measure per-strip pixel shift 4. Strips with consistent shift in one direction = scrolling band 5. Stitch the band across frames into a single wide/tall panorama image 6. OCR the panorama using a vision language model (Gemini) 7. Deduplicate repeated marquee loops (e.g. "SALE • SALE • SALE •" → "SALE") **Performance**: - Processing speed: ~2-5x realtime (depends on video resolution and FPS) - Accuracy: Best with consistent scroll speed; handles variable speed with degradation - Minimum video length: ~2 seconds (needs 3+ frames for correlation) **Supported Scroll Directions**: - Horizontal: Right-to-left (most common), Left-to-right - Vertical: Bottom-to-top (credits), Top-to-bottom examples: - description: Standard video ad with scrolling banner extractor_type: scrolling_text_extractor fps: 5 strip_height: 40 use_case: Extract promotional text from video ad tickers - description: Fast-scrolling news ticker extractor_type: scrolling_text_extractor fps: 10 min_shift_px: 3 strip_height: 30 use_case: Capture rapidly scrolling chyron text from news broadcasts - consistency_ratio: 0.5 description: Credits / disclaimer scroll extractor_type: scrolling_text_extractor fps: 5 min_shift_px: 1.5 strip_height: 60 use_case: Extract vertically scrolling credits or legal disclaimers TextExtractorParams: properties: extractor_type: type: string const: text_extractor title: Extractor Type description: Discriminator field for parameter type identification. default: text_extractor source_type: type: string enum: - text - youtube title: Source Type description: >- Source content type. Use 'youtube' to resolve YouTube URLs to caption text before embedding. Default: 'text' (plain text input). default: text split_by: $ref: '#/components/schemas/TextSplitStrategy' description: Strategy for splitting text into multiple documents. default: none chunk_size: type: integer maximum: 10000 minimum: 1 title: Chunk Size description: Target size for each chunk. default: 1000 chunk_overlap: type: integer maximum: 5000 minimum: 0 title: Chunk Overlap description: Number of units to overlap between consecutive chunks. default: 0 segment_length_seconds: type: integer maximum: 600 minimum: 30 title: Segment Length Seconds description: >- Length of each transcript segment in seconds (for time_segments split strategy). Shorter segments give more precise search results but more documents. default: 120 language: type: string title: Language description: >- Preferred language code for YouTube captions (when source_type='youtube'). default: en extract_captions: type: boolean title: Extract Captions description: >- Extract auto-captions or manual subtitles from YouTube videos (when source_type='youtube'). Falls back to video description if False. default: true response_shape: anyOf: - type: string - additionalProperties: true type: object - type: 'null' title: Response Shape description: Define custom structured output using LLM extraction. llm_provider: anyOf: - type: string - type: 'null' title: Llm Provider description: LLM provider for structured extraction (openai, google, anthropic). llm_model: anyOf: - type: string - type: 'null' title: Llm Model description: Specific LLM model for structured extraction. llm_api_key: anyOf: - type: string - type: 'null' title: Llm Api Key description: >- API key for LLM operations (BYOK - Bring Your Own Key). Supports: - Direct key: 'sk-proj-abc123...' - Secret reference: '{{SECRET.openai_api_key}}' When using secret reference, the key is loaded from your organization's secrets vault at runtime. Store secrets via POST /v1/organizations/secrets. If not provided, uses Mixpeek's default API keys. embedding_model: anyOf: - $ref: '#/components/schemas/EmbeddingModel' - type: 'null' description: >- Embedding model to use. Defaults to the current TEXT modality default in the central embedding registry. Changing this on an existing namespace requires a migration — dimensions are fixed at namespace creation. embedding_task: anyOf: - type: string - type: 'null' title: Embedding Task description: >- Embedding task hint for instruction-aware models (E5, Gemini). Prefer setting this at collection level (embedding_task on the collection) rather than here. Collection-level overrides this value. Defaults to 'retrieval_document'. Values: retrieval_document, retrieval_query, semantic_similarity, classification, clustering. type: object title: TextExtractorParams description: >- Parameters for the text extractor. The text extractor generates dense vector embeddings optimized for semantic similarity search. It uses the E5-Large multilingual model to convert text into 1024-dimensional vectors. When ``source_type`` is ``"youtube"``, the extractor first resolves YouTube URLs to caption text via yt-dlp before chunking and embedding. Use ``split_by="time_segments"`` with ``segment_length_seconds`` to segment captions by time window. examples: - chunk_overlap: 0 chunk_size: 1000 extractor_type: text_extractor split_by: none - chunk_overlap: 1 chunk_size: 5 extractor_type: text_extractor split_by: sentences - extractor_type: text_extractor language: en segment_length_seconds: 120 source_type: youtube split_by: time_segments UniversalExtractorParams: properties: extractor_type: type: string const: universal_extractor title: Extractor Type description: Discriminator field for parameter type identification. default: universal_extractor output_dimensionality: type: integer maximum: 3072 minimum: 256 title: Output Dimensionality description: Output embedding dimensions (Gemini Embedding 2 supports 256-3072). default: 3072 task_type: type: string title: Task Type description: >- Embedding intent used as a text instruction for Gemini Embedding 2. Common values: RETRIEVAL_DOCUMENT, RETRIEVAL_QUERY, SEMANTIC_SIMILARITY. default: RETRIEVAL_DOCUMENT generate_description: type: boolean title: Generate Description description: >- Generate a text description of the content via Gemini vision/understanding. default: true extract_text: type: boolean title: Extract Text description: >- Extract text content (OCR for images/docs, transcription for audio/video). default: true max_video_segments: type: integer maximum: 50 minimum: 1 title: Max Video Segments description: Maximum number of 30s segments to process for video files. default: 10 max_document_pages: type: integer maximum: 200 minimum: 1 title: Max Document Pages description: Maximum number of pages to process for document files. default: 50 max_file_download_mb: type: integer maximum: 1024 minimum: 1 title: Max File Download Mb description: Maximum file download size in MB for Celery fast-path processing. default: 500 max_concurrency: type: integer maximum: 32 minimum: 1 title: Max Concurrency description: Maximum per-task object concurrency for Celery fast-path processing. default: 4 type: object title: UniversalExtractorParams description: Parameters for the Universal Extractor. examples: - extract_text: true extractor_type: universal_extractor generate_description: true max_concurrency: 4 max_file_download_mb: 500 output_dimensionality: 3072 task_type: RETRIEVAL_DOCUMENT WebScraperExtractorParams: properties: extractor_type: type: string const: web_scraper title: Extractor Type description: Discriminator field for parameter type identification. default: web_scraper max_depth: type: integer maximum: 10 minimum: 0 title: Max Depth description: >- Maximum link depth to crawl. 0=seed page only, 1=seed+direct links, etc. Default: 2. Max: 10. default: 2 max_pages: type: integer maximum: 500 minimum: 1 title: Max Pages description: 'Maximum pages to crawl. Default: 50. Max: 500.' default: 50 crawl_timeout: type: integer maximum: 3600 minimum: 10 title: Crawl Timeout description: >- Maximum total time for crawling in seconds. Default: 300 (5 minutes). Increase for large sites with many pages. Max: 3600 (1 hour). default: 300 crawl_mode: $ref: '#/components/schemas/CrawlMode' description: >- Crawl strategy. DETERMINISTIC: BFS all links (predictable). SEMANTIC: LLM-guided, prioritizes relevant pages (requires crawl_goal). default: deterministic crawl_goal: anyOf: - type: string - type: 'null' title: Crawl Goal description: >- Goal for semantic crawling. Only used when crawl_mode=SEMANTIC. Example: 'Find all S3 API documentation and examples' render_strategy: $ref: '#/components/schemas/RenderStrategy' description: >- How to render pages. AUTO (default): tries static, falls back to JS. STATIC: fast HTTP fetch. JAVASCRIPT: Playwright browser for SPAs. default: auto include_patterns: anyOf: - items: type: string type: array - type: 'null' title: Include Patterns description: 'Regex patterns for URLs to include. Example: [''/docs/'', ''/api/'']' exclude_patterns: anyOf: - items: type: string type: array - type: 'null' title: Exclude Patterns description: 'Regex patterns for URLs to exclude. Example: [''/blog/'', ''\.pdf$'']' chunk_strategy: $ref: '#/components/schemas/ChunkStrategy' description: >- How to split page content. NONE: one chunk per page. SENTENCES/PARAGRAPHS: semantic boundaries. WORDS/CHARACTERS: fixed size chunks. default: none chunk_size: type: integer maximum: 10000 minimum: 1 title: Chunk Size description: Target size for each chunk (in units of chunk_strategy). default: 500 chunk_overlap: type: integer maximum: 5000 minimum: 0 title: Chunk Overlap description: Overlap between chunks to preserve context. default: 50 document_id_strategy: $ref: '#/components/schemas/DocumentIdStrategy' description: >- How to generate document IDs. URL (default): stable across re-crawls. POSITION: order-based. CONTENT: deduplicates identical content. default: url generate_text_embeddings: type: boolean title: Generate Text Embeddings description: Generate E5 embeddings for text content. default: true generate_code_embeddings: type: boolean title: Generate Code Embeddings description: Generate Jina code embeddings for code blocks. default: true generate_image_embeddings: type: boolean title: Generate Image Embeddings description: Generate SigLIP embeddings for images/figures. default: true generate_structure_embeddings: type: boolean title: Generate Structure Embeddings description: Generate DINOv2 visual structure embeddings for layout comparison. default: true response_shape: anyOf: - type: string - additionalProperties: true type: object - type: 'null' title: Response Shape description: >- Optional structured extraction schema. Natural language or JSON schema. Example: 'Extract API version, deprecated methods, and example code' llm_provider: anyOf: - type: string - type: 'null' title: Llm Provider description: 'LLM provider for structured extraction: openai, google, anthropic' llm_model: anyOf: - type: string - type: 'null' title: Llm Model description: LLM model for structured extraction. llm_api_key: anyOf: - type: string - type: 'null' title: Llm Api Key description: >- API key for LLM operations (BYOK - Bring Your Own Key). Supports: - Direct key: 'sk-proj-abc123...' - Secret reference: '{{SECRET.openai_api_key}}' When using secret reference, the key is loaded from your organization's secrets vault at runtime. Store secrets via POST /v1/organizations/secrets. If not provided, uses Mixpeek's default API keys. max_retries: type: integer maximum: 10 minimum: 0 title: Max Retries description: >- Maximum retry attempts for failed HTTP requests. Uses exponential backoff with jitter. Default: 3. default: 3 retry_base_delay: type: number maximum: 30 minimum: 0.1 title: Retry Base Delay description: >- Base delay in seconds for retry backoff. Actual delay = base * 2^attempt + jitter. Default: 1.0. default: 1 retry_max_delay: type: number maximum: 300 minimum: 1 title: Retry Max Delay description: 'Maximum delay in seconds between retries. Default: 30.' default: 30 respect_retry_after: type: boolean title: Respect Retry After description: >- Respect Retry-After header from 429/503 responses. If False, uses exponential backoff instead. Default: True. default: true proxies: anyOf: - items: type: string type: array - type: 'null' title: Proxies description: >- List of proxy URLs for rotation. Supports formats: 'http://host:port', 'http://user:pass@host:port', 'socks5://host:port'. Proxies rotate on errors or every N requests. rotate_proxy_on_error: type: boolean title: Rotate Proxy On Error description: 'Rotate to next proxy when request fails. Default: True.' default: true rotate_proxy_every_n_requests: type: integer maximum: 1000 minimum: 0 title: Rotate Proxy Every N Requests description: >- Rotate proxy every N requests (0 = disabled). Useful for avoiding IP-based rate limits. Default: 0 (disabled). default: 0 captcha_service_provider: anyOf: - type: string - type: 'null' title: Captcha Service Provider description: >- Captcha solving service provider: '2captcha', 'anti-captcha', 'capsolver'. If not set, captcha pages are skipped gracefully. captcha_service_api_key: anyOf: - type: string - type: 'null' title: Captcha Service Api Key description: >- API key for captcha solving service. Supports secret reference: '{{SECRET.captcha_api_key}}'. Required if captcha_service_provider is set. detect_captcha: type: boolean title: Detect Captcha description: >- Detect captcha challenges (Cloudflare, reCAPTCHA, hCaptcha). If detected and no solver configured, page is skipped. Default: True. default: true persist_cookies: type: boolean title: Persist Cookies description: >- Persist cookies across requests within a crawl session. Useful for sites requiring authentication. Default: True. default: true custom_headers: anyOf: - additionalProperties: type: string type: object - type: 'null' title: Custom Headers description: >- Custom HTTP headers to include in all requests. Example: {'Authorization': 'Bearer token', 'X-Custom': 'value'} youtube_mode: type: string enum: - auto - 'off' - force title: Youtube Mode description: >- YouTube channel fast path. 'auto' (default): detect YouTube channel URLs and hand off to yt-dlp; other URLs run normal BFS. 'off': never hand off. 'force': treat every URL as a YouTube channel (useful for explicit channel buckets). default: auto youtube_max_videos: type: integer maximum: 5000 minimum: 1 title: Youtube Max Videos description: 'Max videos to pull per channel enumeration. Default: 50.' default: 50 youtube_backfill_months: type: integer maximum: 60 minimum: 1 title: Youtube Backfill Months description: 'Skip videos older than this many months. Default: 6.' default: 6 youtube_show_filter: anyOf: - type: string - type: 'null' title: Youtube Show Filter description: >- Optional case-insensitive regex applied to video titles. Used when a channel hosts multiple shows and you only want one. youtube_download_videos: type: boolean title: Youtube Download Videos description: >- If true (default), download each video file so downstream processors can read it from the `video_path` field. If false, only metadata + captions are emitted. default: true youtube_format_ladder: type: string title: Youtube Format Ladder description: >- yt-dlp format selector. Defaults to a muxed-first ladder that avoids breaking on YouTube's DASH n-challenge when the current EJS signature solver is stale. See BUILD_LOG.md in the greenroom folder for the rationale. default: b[height<=720]/bv*[height<=720]+ba/best youtube_cookies_path: anyOf: - type: string - type: 'null' title: Youtube Cookies Path description: >- Optional path to a Netscape cookies file. Required for age-gated or members-only content. youtube_request_sleep: type: number maximum: 30 minimum: 0 title: Youtube Request Sleep description: 'Seconds to sleep between yt-dlp requests. Default: 1.0.' default: 1 delay_between_requests: type: number maximum: 60 minimum: 0 title: Delay Between Requests description: >- Delay in seconds between consecutive requests. Useful for polite crawling and avoiding rate limits. Default: 0 (no delay). default: 0 type: object title: WebScraperExtractorParams description: >- Parameters for the web scraper extractor. The web scraper extractor crawls websites and extracts content with three types of embeddings for comprehensive multimodal search: **Embedding Types:** - Text (E5-Large): 1024D embeddings for page content - Code (Jina Code): 768D embeddings for code blocks - Images (SigLIP): 768D semantic embeddings for figures/screenshots - Images (DINOv2): 768D structure embeddings for visual layout comparison **Crawl Modes:** - DETERMINISTIC: BFS following all links (default, predictable) - SEMANTIC: LLM-guided, prioritizes pages matching crawl_goal **Rendering Strategies:** - STATIC: Fast HTTP fetch (default, works for most sites) - JAVASCRIPT: Playwright browser for SPAs (React/Vue/Angular) - AUTO: Tries static, falls back to JS if content too short **Use Cases:** - Documentation freshness: Crawl docs, compare against course content - Job board ingestion: Extract job listings with structured data - Knowledge base building: Convert websites to searchable collections - Code example indexing: Find API usage patterns across docs examples: - chunk_size: 3 chunk_strategy: paragraphs description: Documentation site crawl extractor_type: web_scraper max_depth: 3 max_pages: 100 - description: Job board extraction extractor_type: web_scraper max_depth: 1 max_pages: 50 render_strategy: auto response_shape: Extract job title, department, location, and requirements - crawl_goal: Find all S3 upload examples and API documentation crawl_mode: semantic description: Semantic crawl for API docs extractor_type: web_scraper generate_code_embeddings: true max_pages: 200 - delay_between_requests: 0.5 description: Large-scale catalogue with resilience extractor_type: web_scraper max_depth: 5 max_pages: 10000 max_retries: 5 respect_retry_after: true - description: Protected site with proxy rotation extractor_type: web_scraper max_pages: 5000 persist_cookies: true proxies: - http://proxy1.example.com:8080 - http://proxy2.example.com:8080 rotate_proxy_every_n_requests: 50 rotate_proxy_on_error: true CustomPluginParams: properties: extractor_type: type: string title: Extractor Type description: Custom plugin extractor type (plugin name) additionalProperties: true type: object required: - extractor_type title: CustomPluginParams description: >- Parameters for custom plugin extractors. This model accepts any extractor_type that doesn't match the builtin extractors, allowing custom plugins to define their own parameters. InputMapping: properties: input_key: type: string title: Input Key description: Key used in the constructed inputs payload. source_type: anyOf: - $ref: '#/components/schemas/InputSourceType' - type: 'null' description: Source of the value (payload, literal, vector). path: anyOf: - type: string - type: 'null' title: Path description: >- Dot-notation path inside payload/vector when source_type is PAYLOAD or VECTOR. override: anyOf: - {} - type: 'null' title: Override description: Static value used when source_type is LITERAL. Overrides any path. type: object required: - input_key title: InputMapping description: |- Declarative mapping for building inputs from various sources. - input_key: The key used in the constructed inputs payload - source_type: Where to fetch the value (payload, literal, vector) - path: Dot-notation path when source_type is PAYLOAD or VECTOR - override: Static value when source_type is LITERAL examples: - input_key: query_text path: content.title source_type: payload - input_key: lang override: en source_type: literal - input_key: image_vector path: features.clip_vit_l_14 source_type: vector FieldPassthrough: properties: source_path: type: string title: Source Path description: >- REQUIRED. Path to the source field to copy. Simple fields: Use field name directly (e.g., 'title', 'campaign_id'). Nested fields: Use dot notation (e.g., 'metadata.author', 'config.model.version'). The field must exist in the source bucket schema or upstream collection schema. Without target_path, nested fields are flattened: 'metadata.author' becomes 'author' in output. examples: - title - campaign_id - metadata.author - config.model_version - category target_path: anyOf: - type: string - type: 'null' title: Target Path description: >- OPTIONAL. Target field name in output document. If NOT PROVIDED: Uses source_path name (or last component for nested paths). - 'title' → 'title' - 'metadata.author' → 'author' If PROVIDED: Uses this exact name in output. - source_path='doc_title', target_path='title' → 'title' - source_path='metadata.author', target_path='contributor' → 'contributor' Use cases: - Rename fields for cleaner API schemas - Avoid name conflicts with extractor outputs - Standardize field names across different sources Constraints: - Must not conflict with system fields (document_id, collection_id, etc.) - Must not conflict with extractor output fields - Must be a valid field name (alphanumeric, underscores, hyphens) examples: - title - author - campaign_name - product_id - user_email default: anyOf: - {} - type: 'null' title: Default description: >- OPTIONAL. Default value if source field doesn't exist or is None. If NOT PROVIDED and field missing: Field is omitted from output document. If PROVIDED and field missing: Field is included with this default value. Type should match expected field type (string, int, list, dict, etc.). examples: - Unknown - N/A - 0 - [] - {} - false required: type: boolean title: Required description: >- OPTIONAL. Whether this field MUST exist in source. If True and field missing: Raises validation error, processing fails. If False and field missing: Field omitted (or default used if provided). Use True for: Critical identifiers, required business fields. Use False for: Optional metadata, nice-to-have fields. Default: False (field is optional). default: false type: object required: - source_path title: FieldPassthrough description: >- Configuration for passing fields from source to output documents. Simple field passthrough: specify which fields to copy from source (bucket object or upstream collection document) to the output documents alongside extractor outputs. Use Cases: - Preserve identifiers: campaign_id, product_sku, user_id - Keep metadata: category, tags, author, timestamp - Maintain business context: priority, status, region - Extract nested values: metadata.author, config.model_version - Rename fields for cleaner schemas: doc_title → title Field Selection: - WITHOUT field_passthrough: Only extractor outputs appear in documents - WITH field_passthrough: Specified fields + extractor outputs - WITH include_all_source_fields=True: All source fields + extractor outputs Field Naming: - WITHOUT target_path: Output uses source name (or last component for nested) - "title" → "title" - "metadata.author" → "author" - WITH target_path: Output uses specified name - source_path="doc_title", target_path="title" → "title" - source_path="metadata.author", target_path="contributor" → "contributor" Requirements: - source_path is REQUIRED - specifies which field to copy (supports dot notation) - target_path is OPTIONAL - rename field in output (default: auto-derived name) - default is OPTIONAL - provides fallback if field missing (default: omit field) - required is OPTIONAL - errors if field missing (default: false, omit field) examples: - description: Simple field passthrough required: true source_path: title - default: uncategorized description: Field with default value source_path: category - description: Field renaming for cleaner schema source_path: doc_title target_path: title - description: Nested field extraction with custom name source_path: metadata.author target_path: contributor - description: Required field with renaming required: true source_path: campaign_id target_path: campaign - default: normal description: Optional field with default and renaming required: false source_path: metadata.priority target_path: priority SourceType: type: string enum: - bucket - collection - taxonomy - cluster - direct_upsert - none title: SourceType description: Source types for any document/point. SourceFilters-Output: properties: filters: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - type: 'null' description: >- Optional logical filters to apply to source data. Uses LogicalOperator model with AND/OR/NOT support. When specified, only objects/documents matching these filters will be processed by this collection. When null, all source data is processed (no filtering). Filters are consistent across all batch runs for this collection. examples: - AND: - field: blobs.type operator: eq value: video - AND: - field: metadata.status operator: eq value: active - field: created_at operator: gte value: '2025-10-01T00:00:00Z' - OR: - field: brand_name operator: in value: - Acme - TechCo - field: category operator: eq value: premium type: object title: SourceFilters description: >- Filters applied to source data when processing collections. Source filters determine which objects (from buckets) or documents (from collections) are processed by this collection. Filters use the same LogicalOperator model as list APIs throughout the system, supporting complex AND/OR/NOT logic. Use Cases: - Process only specific content types from mixed-content buckets - Filter by metadata fields (status, category, tags, dates) - Create specialized collections from broader sources - Exclude certain objects or documents from processing Examples: Process only video content: { "AND": [ {"field": "blobs.type", "operator": "eq", "value": "video"} ] } Process only active, published content: { "AND": [ {"field": "metadata.status", "operator": "eq", "value": "active"}, {"field": "metadata.published", "operator": "eq", "value": true} ] } Process content from last 30 days: { "AND": [ {"field": "created_at", "operator": "gte", "value": "2025-10-08T00:00:00Z"} ] } Process specific brands OR categories: { "OR": [ {"field": "brand_name", "operator": "in", "value": ["Acme", "TechCo"]}, {"field": "category", "operator": "eq", "value": "premium"} ] } Filter Operators: - eq (equals) - ne (not equals) - gt (greater than) - gte (greater than or equal) - lt (less than) - lte (less than or equal) - in (value in list) - nin (value not in list) - contains (string contains) - starts_with (string starts with) - ends_with (string ends with) Performance Considerations: - Filters are evaluated at batch creation time - Only matching objects/documents are included in processing - More selective filters = smaller batches = faster processing - Use indexed fields (metadata, timestamps) for better performance Relationship to Batch Filters: - Source filters: Applied at collection definition (consistent across all batches) - Batch filters: Applied at batch creation (ad-hoc, per-batch basis) - Both can be used together: source filters + batch filters = intersection examples: - description: Filter only video content filters: AND: - field: blobs.type operator: eq value: video - description: Filter active content from specific brands filters: AND: - field: metadata.status operator: eq value: active - field: brand_name operator: in value: - Acme - TechCo - Innovate - description: Filter by content type and date range filters: AND: - OR: - field: blobs.type operator: eq value: video - field: blobs.type operator: eq value: image - field: created_at operator: gte value: '2025-10-01T00:00:00Z' - field: created_at operator: lte value: '2025-10-31T23:59:59Z' - description: No filters - process all source data TaxonomyExecutionMode: type: string enum: - materialize title: TaxonomyExecutionMode description: How a taxonomy should be executed when attached to a collection. LogicalOperator-Output: properties: AND: anyOf: - items: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - $ref: '#/components/schemas/FilterCondition' type: array - type: 'null' title: And description: Logical AND operation - all conditions must be true example: - field: name operator: eq value: John - field: age operator: gte value: 30 OR: anyOf: - items: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - $ref: '#/components/schemas/FilterCondition' type: array - type: 'null' title: Or description: Logical OR operation - at least one condition must be true example: - field: status operator: eq value: active - field: role operator: eq value: admin NOT: anyOf: - items: anyOf: - $ref: '#/components/schemas/LogicalOperator-Output' - $ref: '#/components/schemas/FilterCondition' type: array - type: 'null' title: Not description: Logical NOT operation - all conditions must be false example: - field: department operator: eq value: HR - field: location operator: eq value: remote case_sensitive: anyOf: - type: boolean - type: 'null' title: Case Sensitive description: Whether to perform case-sensitive matching default: false example: true additionalProperties: true type: object title: LogicalOperator description: >- Represents a logical operation (AND, OR, NOT) on filter conditions. Allows nesting with a defined depth limit. Also supports shorthand syntax where field names can be passed directly as key-value pairs for equality filtering (e.g., {"metadata.title": "value"}). HierarchicalEnrichmentStyle: type: string enum: - full_chain - best_match - combined title: HierarchicalEnrichmentStyle description: >- How hierarchical taxonomy results should be structured in enriched documents. Controls the field naming pattern for multi-tier taxonomy enrichment. AlertExecutionMode: type: string enum: - on_ingest - scheduled - on_demand title: AlertExecutionMode description: When the alert should execute. AlertInputMapping: properties: input_key: type: string title: Input Key description: The retriever input parameter name examples: - query_embedding - collection_id - category_filter source: $ref: '#/components/schemas/InputMappingSource' description: Where to get the value from type: object required: - input_key - source title: AlertInputMapping description: |- Maps a retriever input to a document field or constant. Input mappings define how to construct retriever inputs from the ingested document. This allows the same alert to be used with different collections that have different field structures. Examples: # Map document embedding to retriever query { "input_key": "query_embedding", "source": {"source_type": "document_field", "path": "features.video_embedding"} } # Use constant collection for search target { "input_key": "collection_id", "source": {"source_type": "constant", "value": "col_known_incidents"} } Attributes: input_key: The retriever input parameter name source: Where to get the value from examples: - input_key: query_embedding source: path: features.video_embedding source_type: document_field - input_key: target_collection source: source_type: constant value: col_known_incidents PostProcessingPhase: type: integer enum: - 1 - 2 - 3 - 4 title: PostProcessingPhase description: >- Execution phases for post-processing applications. Applications execute in phase order (lower = earlier). Within a phase, applications execute by priority (higher = earlier). Phases: TAXONOMY (1): Classification and labeling operations CLUSTER (2): Grouping and clustering operations ALERT (3): Notifications and alerts (default for alerts) RETRIEVER_ENRICHMENT (4): Retriever-based enrichment operations The default phase for each application type: - TaxonomyApplicationConfig: TAXONOMY - ClusterApplicationConfig: CLUSTER - AlertApplicationConfig: ALERT - RetrieverEnrichmentConfig: RETRIEVER_ENRICHMENT Users can override the phase via the `execution_phase` field to run applications in non-default order. For example, an alert can be configured to run in Phase 1 alongside taxonomies if early notification is needed. Example: # Default: Alert runs after taxonomies and clusters AlertApplicationConfig(alert_id="alt_123", execution_phase=PostProcessingPhase.ALERT) # Override: Run alert early, in taxonomy phase AlertApplicationConfig(alert_id="alt_urgent", execution_phase=PostProcessingPhase.TAXONOMY) EnrichmentInputMapping: properties: input_key: type: string title: Input Key description: The retriever input parameter name examples: - query - query_embedding - collection_id source: $ref: '#/components/schemas/InputMappingSource' description: Where to get the value from (document field or constant) type: object required: - input_key - source title: EnrichmentInputMapping description: >- Maps a document field or constant to a retriever input parameter. Defines how to construct retriever inputs from the document being enriched. Attributes: input_key: The retriever input parameter name source: Where to get the value from (document field or constant) examples: - input_key: query source: path: title source_type: document_field - input_key: collection_id source: source_type: constant value: col_reference_data WriteBackFieldMapping: properties: source_field: type: string title: Source Field description: Field path in retriever result (dot notation supported) examples: - ai_safety_insight.text - category - score target_field: type: string title: Target Field description: Field name to write on the document examples: - _enrichment_category - safety_score - related_items mode: type: string enum: - first - all_as_array - concat title: Mode description: >- How to aggregate values from multiple results. 'first': top result only. 'all_as_array': collect into list. 'concat': join strings with ', '. default: first type: object required: - source_field - target_field title: WriteBackFieldMapping description: |- Maps a field from retriever results back to the document. Controls how retriever result fields are written to the source document. Attributes: source_field: Field path in retriever result (dot notation for nested fields) target_field: Field name to write on the document mode: How to aggregate values across multiple results: - "first": Write value from first result only (default) - "all_as_array": Collect values from all results into a list - "concat": Concatenate string values with ", " separator BucketSchemaFieldType: type: string enum: - string - number - integer - float - boolean - object - array - date - datetime - text - image - audio - video - pdf - excel title: BucketSchemaFieldType description: >- Supported data types for bucket schema fields. Types fall into two categories: 1. **Metadata Types** (JSON types): - Stored as object metadata - Standard JSON-compatible types - Not processed by extractors (unless explicitly mapped) - Examples: string, number, boolean, date 2. **File Types** (blobs): - Stored as files/blobs - Processed by extractors - Require file content (URL or base64) - Examples: text, image, video, pdf **GIF Special Handling**: GIF files can be declared as either IMAGE or VIDEO type: - As IMAGE: GIF is embedded as a single static image (first frame) - As VIDEO: GIF is decomposed frame-by-frame with embeddings per frame The multimodal extractor detects GIFs via MIME type (image/gif) and routes them based on your schema declaration. Use VIDEO for animated GIFs where frame-level search is needed, IMAGE for static/thumbnail use cases. NOTE: For retriever input schemas that need to accept document references (e.g., "find similar documents"), use RetrieverInputSchemaFieldType instead, which includes all bucket types plus document_reference. SplitMethod: type: string enum: - time - scene - silence title: SplitMethod GenerationConfig: properties: candidate_count: type: integer title: Candidate Count description: Number of candidate responses to generate for video description. default: 1 max_output_tokens: type: integer title: Max Output Tokens description: Maximum number of tokens for the generated video description. default: 1024 temperature: type: number title: Temperature description: >- Controls randomness for video description generation. Higher is more random. default: 0.2 top_p: type: number title: Top P description: Nucleus sampling (top-p) for video description generation. default: 0.8 response_mime_type: anyOf: - type: string - type: 'null' title: Response Mime Type description: MIME type for response (e.g., 'application/json') response_schema: anyOf: - additionalProperties: true type: object - type: 'null' title: Response Schema description: JSON schema for structured output type: object title: GenerationConfig description: Configuration for generative models. TextSplitStrategy: type: string enum: - characters - words - sentences - paragraphs - pages - time_segments - none title: TextSplitStrategy description: Strategy for splitting text into chunks. EmbeddingModel: type: string enum: - laion_clip_vit_l_14_v1 - multilingual_e5_large_instruct_v1 - vertex_multimodal_embedding - multimodalembedding@001 - gemini-embedding-2 - google_siglip_base_v1 - google_siglip_so400m_v1 - text-embedding-3-small - text-embedding-3-large - face_identity_arcface_r100_v1 - all_minilm_l6_v2_v1 title: EmbeddingModel description: |- Embedding model identifiers. Format: {provider}_{model_name}_{version} CrawlMode: type: string enum: - deterministic - semantic title: CrawlMode description: |- Mode for crawling web pages. Values: DETERMINISTIC: BFS crawl following all links up to max_depth SEMANTIC: LLM-guided crawl prioritizing pages relevant to crawl_goal RenderStrategy: type: string enum: - static - javascript - auto title: RenderStrategy description: |- Strategy for rendering web pages. Values: STATIC: Fast HTTP fetch, works for most sites JAVASCRIPT: Browser rendering via Playwright for SPAs AUTO: Try static first, fall back to JS if content too short ChunkStrategy: type: string enum: - none - sentences - paragraphs - words - characters title: ChunkStrategy description: Strategy for splitting page content into chunks. DocumentIdStrategy: type: string enum: - url - position - content title: DocumentIdStrategy description: |- Strategy for generating deterministic document IDs. Values: URL: hash(page_url + chunk_index) - stable across re-crawls POSITION: hash(seed_url + page_index + chunk_index) - order-based CONTENT: hash(content) - deduplicates identical content InputSourceType: type: string enum: - payload - literal - vector - blob title: InputSourceType description: Where the value for an input should be retrieved from. FilterCondition: properties: field: type: string title: Field description: Field name to filter on operator: $ref: '#/components/schemas/FilterOperator' description: Comparison operator default: eq value: anyOf: - $ref: '#/components/schemas/DynamicValue' - {} title: Value description: Value to compare against type: object required: - field - value title: FilterCondition description: |- Represents a single filter condition. Attributes: field: The field to filter on operator: The comparison operator value: The value to compare against InputMappingSource: properties: source_type: type: string enum: - document_field - constant - source_blob title: Source Type description: Where the value comes from path: anyOf: - type: string - type: 'null' title: Path description: >- JSONPath to document field (when source_type='document_field'), or blob property name (when source_type='source_blob', e.g. 'image') examples: - features.video_embedding - metadata.category - _taxonomy_product_label - image value: anyOf: - {} - type: 'null' title: Value description: Constant value to use (when source_type='constant') examples: - col_known_incidents - 0.85 - true type: object required: - source_type title: InputMappingSource description: >- Defines how to get a value for a retriever input. Can be a document field reference, a constant value, or a source blob URL. Attributes: source_type: Where the value comes from path: JSONPath to document field (when source_type='document_field'), or blob property name (when source_type='source_blob', e.g. 'image') value: Constant value to use (when source_type='constant') FilterOperator: type: string enum: - eq - ne - gt - lt - gte - lte - in - nin - contains - starts_with - ends_with - regex - exists - is_null - text - phrase title: FilterOperator description: Supported filter operators across database implementations. DynamicValue: properties: type: type: string const: dynamic title: Type default: dynamic field: type: string title: Field description: >- The dot-notation path to the value in the runtime query request, e.g., 'inputs.user_id' examples: - inputs.query_text - filters.AND[0].value type: object required: - field title: DynamicValue description: A value that should be dynamically resolved from the query request. ````