> ## Documentation Index > Fetch the complete documentation index at: https://docs.platform.toloka.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Replace the pipeline graph > Send the complete graph, up to 2 MiB, and optionally `expectedVersion`. Returns the updated pipeline metadata. Invalid configuration or missing secrets returns 400; failed checks and lint findings do not block this write. An attached dataset cannot be removed. A stale version, concurrent edit, or active run returns 409; read the graph again before retrying. ## OpenAPI ````yaml /api-reference/openapi.json put /pipelines/{pipelineId}/graph openapi: 3.1.0 info: description: >- Manage projects, datasets, labeling pipelines, and model training. **Beta:** endpoints, request and response shapes, and error contracts can change without notice or a backward-compatibility promise. title: Toloka Platform API version: 2.0.0 servers: - url: https://platform.toloka.ai/api/v2-beta security: - apiKeyAuth: [] - scopedTokenAuth: [] tags: - description: The authenticated API key owner. name: Account - description: A project is the container for pipelines, datasets, and files. name: Projects - description: Rows and typed fields. Field schemas are immutable once created. name: Datasets - description: Append rows to a dataset from a file. name: Dataset uploads - description: Read a dataset out as a file. name: Dataset exports - description: A pipeline is a graph of nodes that processes each dataset item. name: Pipelines - description: The nodes, connections, fields, and version of a pipeline graph. name: Pipeline graph - description: Execute a pipeline over a dataset and inspect each item. name: Runs - description: >- Only for teams with customer review enabled. Customer review of generation tasks: the team accepts or rejects every task of a node that uses customer review. name: Reviews - description: Per-node checks that a pipeline must pass before deployment. name: Checks - description: Named credentials available to pipeline code. name: Secrets - description: Project files available to pipeline code and labeling interfaces. name: Files - description: >- Global, read-only catalog of the models a synthetic generation node can use, with their budget bounds. name: Generation models - description: >- Global, read-only catalog of the expert audience: the values a node `audience` filter can use. name: Audience - description: Project lists of experts a node can take its workforce from. name: Expert lists - description: >- Experts banned in a project: they cannot be invited and are removed from the work they are in. Available only to teams with expert lists enabled. name: Expert ban list - description: Global, read-only base model catalog. name: Base models - description: Project-owned mutable model-training drafts. name: Trainings - description: Project-owned Models and immutable Model Versions. name: Models - name: Batch inference paths: /pipelines/{pipelineId}/graph: put: tags: - Pipeline graph summary: Replace the pipeline graph description: >- Send the complete graph, up to 2 MiB, and optionally `expectedVersion`. Returns the updated pipeline metadata. Invalid configuration or missing secrets returns 400; failed checks and lint findings do not block this write. An attached dataset cannot be removed. A stale version, concurrent edit, or active run returns 409; read the graph again before retrying. operationId: replacePipelineGraph parameters: - in: path name: pipelineId required: true schema: $ref: '#/components/schemas/PipelineId' requestBody: content: application/json: schema: properties: expectedVersion: $ref: '#/components/schemas/PipelineGraphVersion' graph: $ref: '#/components/schemas/PipelineGraphV1' description: >- Complete pipeline graph in version 1 of the public graph format. required: - graph type: object responses: '200': content: application/json: schema: $ref: '#/components/schemas/Pipeline' description: Pipeline graph replaced headers: RateLimit-Limit: $ref: '#/components/headers/RateLimit-Limit' RateLimit-Policy: $ref: '#/components/headers/RateLimit-Policy' RateLimit-Remaining: $ref: '#/components/headers/RateLimit-Remaining' RateLimit-Reset: $ref: '#/components/headers/RateLimit-Reset' X-Request-Id: $ref: '#/components/headers/X-Request-Id' '400': description: Pipeline graph is too large, invalid, or references missing secrets headers: RateLimit-Limit: $ref: '#/components/headers/RateLimit-Limit' RateLimit-Policy: $ref: '#/components/headers/RateLimit-Policy' RateLimit-Remaining: $ref: '#/components/headers/RateLimit-Remaining' RateLimit-Reset: $ref: '#/components/headers/RateLimit-Reset' X-Request-Id: $ref: '#/components/headers/X-Request-Id' '404': description: Pipeline not found headers: RateLimit-Limit: $ref: '#/components/headers/RateLimit-Limit' RateLimit-Policy: $ref: '#/components/headers/RateLimit-Policy' RateLimit-Remaining: $ref: '#/components/headers/RateLimit-Remaining' RateLimit-Reset: $ref: '#/components/headers/RateLimit-Reset' X-Request-Id: $ref: '#/components/headers/X-Request-Id' '409': description: >- The expected graph version is stale, the graph changed while the request was processed, or the pipeline has an active run and cannot be changed headers: RateLimit-Limit: $ref: '#/components/headers/RateLimit-Limit' RateLimit-Policy: $ref: '#/components/headers/RateLimit-Policy' RateLimit-Remaining: $ref: '#/components/headers/RateLimit-Remaining' RateLimit-Reset: $ref: '#/components/headers/RateLimit-Reset' X-Request-Id: $ref: '#/components/headers/X-Request-Id' '429': description: Rate limit exceeded headers: RateLimit-Limit: $ref: '#/components/headers/RateLimit-Limit' RateLimit-Policy: $ref: '#/components/headers/RateLimit-Policy' RateLimit-Remaining: $ref: '#/components/headers/RateLimit-Remaining' RateLimit-Reset: $ref: '#/components/headers/RateLimit-Reset' X-Request-Id: $ref: '#/components/headers/X-Request-Id' x-codeSamples: - label: TypeScript SDK lang: typescript source: > import assert from "node:assert/strict"; import { createClient } from "@toloka-platform/sdk"; const client = createClient({ baseUrl: "https://platform.toloka.ai", apiKey: process.env.TOLOKA_API_KEY!, }); const snapshot = await client.pipelines.getGraph({ pipelineId: "pl.019a0000-0000-7000-8000-000000000001" }); const graph = structuredClone(snapshot.graph); const root = graph.nodes.find((node) => node.type === "root"); assert(root, "The pipeline has no Start node."); root.settings.itemLimit = 100; const pipeline = await client.pipelines.replaceGraph({ pipelineId: "pl.019a0000-0000-7000-8000-000000000001", graph: graph, expectedVersion: snapshot.version }); console.log(pipeline); components: schemas: PipelineId: description: Pipeline ID with prefix `pl.`. pattern: >- ^pl\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string PipelineGraphVersion: description: >- The snapshot version expected to still be active. A stale version is rejected with HTTP 409. pattern: >- ^ple\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string PipelineGraphV1: additionalProperties: false properties: connections: description: Directed connections between node handles. items: additionalProperties: false description: One directed connection. properties: from: additionalProperties: false description: Node and handle at one end of a connection. properties: handle: description: Input or output handle identifier on that node. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string node: description: Node identifier from `graph.nodes`. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string required: - node - handle type: object settings: additionalProperties: {} description: >- Connection settings; use an empty object for a normal connection. properties: {} type: object to: additionalProperties: false description: Node and handle at one end of a connection. properties: handle: description: Input or output handle identifier on that node. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string node: description: Node identifier from `graph.nodes`. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string required: - node - handle type: object required: - from - to - settings type: object maxItems: 2000 type: array fields: description: Fields used by the pipeline, with names and JSON Schemas. items: additionalProperties: false description: One pipeline field. properties: id: description: Caller-defined field identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string jsonSchema: $ref: '#/components/schemas/JsonSchema' description: JSON Schema defining the accepted values for a field. name: description: Dataset field name bound to this pipeline field. type: string required: - id - name - jsonSchema type: object maxItems: 500 type: array layout: additionalProperties: additionalProperties: items: false maxItems: 2 minItems: 2 prefixItems: - description: X type: number - description: 'Y' type: number type: array maxProperties: 2 propertyNames: pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string type: object description: >- Optional node positions for the visual editor, keyed by node ID and block group; each position is an [x, y] pair. maxProperties: 500 propertyNames: pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string type: object nodes: description: Nodes that make up the pipeline. items: description: One pipeline node. oneOf: - $ref: '#/components/schemas/RootNode' - $ref: '#/components/schemas/CodeNode' - $ref: '#/components/schemas/GenerationNode' - $ref: '#/components/schemas/MatchNode' - $ref: '#/components/schemas/QuorumNode' maxItems: 500 type: array version: const: 1 description: Graph format version; always 1. type: number required: - version - fields - nodes - connections type: object Pipeline: description: A pipeline within a project. properties: createdAt: description: Time the pipeline was created, as an ISO 8601 timestamp. format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z))$ type: string id: $ref: '#/components/schemas/PipelineId' name: description: Name of the pipeline. type: string projectId: $ref: '#/components/schemas/ProjectId' updatedAt: description: Time the pipeline was last updated, as an ISO 8601 timestamp. format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z))$ type: string version: description: >- Pipeline version number; copying a pipeline increments the source version. type: number required: - id - projectId - name - version - createdAt - updatedAt type: object JsonSchema: additionalProperties: false description: JSON Schema defining the accepted values for a field. properties: additionalProperties: anyOf: - description: >- Whether extra object properties are allowed, or the schema they must satisfy; omitted means allowed. type: boolean - $ref: '#/components/schemas/JsonSchema' description: >- Whether extra object properties are allowed, or the schema they must satisfy; omitted means allowed. description: >- Whether extra object properties are allowed, or the schema they must satisfy; omitted means allowed. allOf: description: Combined schemas; a value must satisfy every schema when supplied. items: $ref: '#/components/schemas/JsonSchema' description: JSON Schema defining the accepted values for a field. minItems: 1 type: array anyOf: description: >- Alternative schemas; a value must satisfy at least one when supplied. items: $ref: '#/components/schemas/JsonSchema' description: JSON Schema defining the accepted values for a field. minItems: 1 type: array const: $ref: '#/components/schemas/JsonValue' description: Single required JSON value; omit when no exact value is required. default: $ref: '#/components/schemas/JsonValue' description: Suggested default JSON value; omit when none is specified. description: description: Optional explanation of the values accepted by this schema. type: string enum: description: >- Allowed JSON values; omit to allow any value satisfying the other constraints. items: $ref: '#/components/schemas/JsonValue' minItems: 1 type: array examples: description: Example JSON values for this schema; omit when none are supplied. items: $ref: '#/components/schemas/JsonValue' type: array exclusiveMaximum: description: Exclusive upper bound for a number; omit when none is required. type: number exclusiveMinimum: description: Exclusive lower bound for a number; omit when none is required. type: number format: description: >- String format, such as `date-time`, `email`, or `uri`; omit when no format is required. type: string items: $ref: '#/components/schemas/JsonSchema' description: >- Schema that every array item must satisfy; omit to leave items unrestricted. maxItems: description: Maximum number of array items; omit for no upper bound. maximum: 9007199254740991 minimum: 0 type: integer maxLength: description: >- Maximum string length in Unicode characters; omit for no upper bound. maximum: 9007199254740991 minimum: 0 type: integer maximum: description: Inclusive upper bound for a number; omit for no upper bound. type: number minItems: description: Minimum number of array items; omit for no lower bound. maximum: 9007199254740991 minimum: 0 type: integer minLength: description: >- Minimum string length in Unicode characters; omit for no lower bound. maximum: 9007199254740991 minimum: 0 type: integer minimum: description: Inclusive lower bound for a number; omit for no lower bound. type: number oneOf: description: Alternative schemas; a value must satisfy exactly one when supplied. items: $ref: '#/components/schemas/JsonSchema' description: JSON Schema defining the accepted values for a field. minItems: 1 type: array pattern: description: >- Regular expression a string must match; omit for no pattern constraint. type: string properties: $ref: '#/components/schemas/JsonSchemaProperties' description: Schemas for named object properties; omit when none are specified. required: description: >- Object property names that must be present; omit when none are required. items: type: string type: array title: description: Optional display title for this schema. type: string type: anyOf: - description: Accepted JSON type or types; omit to allow any type. enum: - string - number - integer - boolean - object - array - 'null' type: string - description: Accepted JSON type or types; omit to allow any type. items: enum: - string - number - integer - boolean - object - array - 'null' type: string minItems: 1 type: array description: Accepted JSON type or types; omit to allow any type. uniqueItems: description: >- Whether array items must be unique; omitted means duplicates are allowed. type: boolean type: object RootNode: additionalProperties: false description: One pipeline node. properties: id: description: Caller-defined node identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string settings: $ref: '#/components/schemas/RootNodeSettings' type: const: root description: >- Node type: `root` supplies Start and End, `code` runs Python, `generation` creates answers, `match` routes items, and `quorum` combines expert answers. type: string required: - id - type - settings type: object CodeNode: additionalProperties: false description: One pipeline node. properties: id: description: Caller-defined node identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string settings: $ref: '#/components/schemas/CodeNodeSettings' type: const: code description: >- Node type: `root` supplies Start and End, `code` runs Python, `generation` creates answers, `match` routes items, and `quorum` combines expert answers. type: string required: - id - type - settings type: object GenerationNode: additionalProperties: false description: One pipeline node. properties: id: description: Caller-defined node identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string settings: $ref: '#/components/schemas/GenerationNodeSettings' type: const: generation description: >- Node type: `root` supplies Start and End, `code` runs Python, `generation` creates answers, `match` routes items, and `quorum` combines expert answers. type: string required: - id - type - settings type: object MatchNode: additionalProperties: false description: One pipeline node. properties: id: description: Caller-defined node identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string settings: $ref: '#/components/schemas/MatchNodeSettings' type: const: match description: >- Node type: `root` supplies Start and End, `code` runs Python, `generation` creates answers, `match` routes items, and `quorum` combines expert answers. type: string required: - id - type - settings type: object QuorumNode: additionalProperties: false description: One pipeline node. properties: id: description: Caller-defined node identifier, unique within the graph. pattern: ^(?!(?:__proto__|constructor|prototype)$).+$ type: string settings: $ref: '#/components/schemas/QuorumNodeSettings' type: const: quorum description: >- Node type: `root` supplies Start and End, `code` runs Python, `generation` creates answers, `match` routes items, and `quorum` combines expert answers. type: string required: - id - type - settings type: object ProjectId: description: Project ID with prefix `pp.`. pattern: >- ^pp\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string JsonValue: anyOf: - type: string - type: number - type: boolean - type: 'null' - items: $ref: '#/components/schemas/JsonValue' type: array - additionalProperties: $ref: '#/components/schemas/JsonValue' propertyNames: type: string type: object JsonSchemaProperties: additionalProperties: $ref: '#/components/schemas/JsonSchema' propertyNames: type: string type: object RootNodeSettings: additionalProperties: false description: Configuration for this node type. properties: completedFieldProxies: description: Fields collected by the End node. properties: input: description: Fields written at the pipeline output. items: additionalProperties: false description: Field binding. properties: id: description: Field identifier from `graph.fields`. type: string optional: description: Whether this field may be absent when the node executes. type: boolean required: - id - optional type: object type: array required: - input type: object datasetId: description: >- Attached dataset ID with prefix `ds.`; absent before a dataset is attached. pattern: >- ^ds\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string itemFilter: additionalProperties: false description: >- Conditions selecting dataset rows; omitted conditions do not restrict the selection. properties: createdBefore: description: >- Cutoff for dataset items and field revisions, as an ISO 8601 timestamp; omit to use the endpoint's cutoff. format: date-time pattern: >- ^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d:[0-5]\d(?:\.\d+)?(?:Z))$ type: string excludeParts: $ref: '#/components/schemas/DatasetSplitPartExclusions' fieldSources: $ref: '#/components/schemas/DatasetFieldSourceConditions' fieldValues: $ref: '#/components/schemas/DatasetFieldValueConditions' missingRevisionFields: description: >- Field IDs with prefix `df.` that must have no revision on a selected row; defaults to an empty array. items: pattern: >- ^df\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string type: array parts: $ref: '#/components/schemas/DatasetSplitPartInclusion' slice: $ref: '#/components/schemas/DatasetSliceSelection' type: object itemLimit: description: >- Maximum number of matching items to process; omit to process every matching item. exclusiveMinimum: 0 maximum: 9007199254740991 type: integer startedFieldProxies: description: Fields supplied by the Start node. properties: output: description: Dataset fields read at the pipeline input. items: additionalProperties: false description: Field binding. properties: id: description: Field identifier from `graph.fields`. type: string optional: description: Whether this field may be absent when the node executes. type: boolean required: - id - optional type: object type: array required: - output type: object required: - startedFieldProxies - completedFieldProxies type: object CodeNodeSettings: additionalProperties: false description: Configuration for this node type. properties: about: $ref: '#/components/schemas/NodeAbout' checks: additionalProperties: $ref: '#/components/schemas/CheckTestCases' description: >- Configured test inputs grouped by check type; absent when no tests are configured. propertyNames: enum: - code type: string type: object code: additionalProperties: false description: Code executed by this node. properties: code: description: Python source code for the node. type: string required: - code type: object envVars: default: [] description: >- Environment variables available to the node code; defaults to an empty array. items: description: One environment variable for node code. oneOf: - additionalProperties: false description: One environment variable for node code. properties: name: description: Environment variable name available to the node code. type: string type: const: static description: >- Value source: `static` uses the supplied value; `secret` reads a named project secret. type: string value: description: Literal value for a static environment variable. type: string required: - type - name - value type: object - additionalProperties: false description: One environment variable for node code. properties: name: description: Environment variable name available to the node code. type: string secretName: description: >- Name of the project secret supplying this environment variable. type: string type: const: secret description: >- Value source: `static` uses the supplied value; `secret` reads a named project secret. type: string required: - type - name - secretName type: object type: array fieldProxies: description: Input and output field bindings for the node. properties: input: description: Fields available to this node as input. items: additionalProperties: false description: Field binding. properties: id: description: Field identifier from `graph.fields`. type: string optional: description: Whether this field may be absent when the node executes. type: boolean required: - id - optional type: object type: array output: description: Fields this node produces. items: additionalProperties: false description: Field binding. properties: id: description: Field identifier from `graph.fields`. type: string optional: description: Whether this field may be absent when the node executes. type: boolean required: - id - optional type: object type: array required: - input - output type: object required: - about - fieldProxies - code type: object GenerationNodeSettings: additionalProperties: false description: Configuration for this node type. properties: about: $ref: '#/components/schemas/NodeAbout' audienceAndPricing: description: Who completes tasks and how attempts are priced. oneOf: - $ref: '#/components/schemas/HumanAudienceAndPricing' - additionalProperties: false description: Who completes tasks and how attempts are priced. properties: audienceType: const: synthetic description: >- Task performer: `human` for experts or `synthetic` for model-generated answers. type: string generationModelId: default: gpt-5.6-terra description: Generation model ID from the generation-models endpoint. enum: - gpt-5.4 - gpt-5.4-mini - claude-sonnet-4-6 - gemini-3.1-pro-preview - gemini-3.5-flash - gemini-3.7-flash - claude-sonnet-5 - claude-opus-5 - claude-fable-5 - gpt-5.6-terra - gpt-5.6-luna - gpt-5.6-sol type: string llmInstruction: default: '' description: >- Instruction supplied to the generation model; defaults to an empty string. type: string maxCostPerAttempt: description: Maximum spend for one generation attempt, in US dollars. type: number reasoningEffort: description: >- Reasoning effort for models that support it; omit to use the model default. enum: - low - medium - high type: string required: - audienceType - maxCostPerAttempt type: object checks: additionalProperties: $ref: '#/components/schemas/CheckTestCases' description: >- Configured test inputs grouped by check type; absent when no tests are configured. propertyNames: enum: - ui - llmqa - annotation_qa - qc_lint - ui_types_lint - exam_dataset - synthgen type: string type: object entities: $ref: '#/components/schemas/TaskEntities' exam: $ref: '#/components/schemas/QualificationExam' instruction: description: Task instructions shown to experts. type: string onboarding: $ref: '#/components/schemas/Onboarding' publicName: description: Task name shown to experts. type: string qualityAssurance: description: How answers are checked for quality. oneOf: - description: How answers are checked for quality. properties: llmQaModelId: default: gpt-5.6-terra description: Model identifier for LLM quality checks. enum: - gpt-5.6-terra - gpt-5.6-luna - gpt-5.4 - gpt-5.4-mini - gpt-5 - gpt-5-mini - claude-sonnet-4-6 type: string qaMode: const: llm_only description: >- Review mode: `llm_only` uses LLM checks, `human_only` uses human review, and `hybrid` combines them. type: string required: - qaMode type: object - additionalProperties: false description: How answers are checked for quality. properties: expertListCondition: anyOf: - $ref: '#/components/schemas/ExpertListCondition' - type: 'null' description: >- The expert lists the node takes its experts from: experts in any of the `list_in` lists, but not in any of the `list_nin` lists. One list is `{ operator: "list_in", expertListId }` with an `exl.` id; several are `{ logic: "or", predicates: [list_in…] }`; excluding lists wraps that in `{ logic: "and", predicates: [, { operator: "list_nin", expertListId }…] }`. Other combinations (intersections, nested exclusions) are rejected. The platform stores the condition in canonical form (groups flattened, predicates deduplicated and sorted). Not available to teams on the older team ban list. `null` detaches the lists. Omitted when no lists are attached. humanQaTimeForTask: description: >- Expected human review time per task in seconds; absent when not specified. exclusiveMinimum: 0 type: number qaMode: const: human_only description: >- Review mode: `llm_only` uses LLM checks, `human_only` uses human review, and `hybrid` combines them. type: string smartSamplingPercent: default: 100 description: >- Percentage of answers sampled for human review, from 10 to 100; defaults to 100 for human-only review and 30 for hybrid review. maximum: 100 minimum: 10 type: number required: - qaMode type: object - additionalProperties: false description: How answers are checked for quality. properties: expertListCondition: anyOf: - $ref: '#/components/schemas/ExpertListCondition' - type: 'null' description: >- The expert lists the node takes its experts from: experts in any of the `list_in` lists, but not in any of the `list_nin` lists. One list is `{ operator: "list_in", expertListId }` with an `exl.` id; several are `{ logic: "or", predicates: [list_in…] }`; excluding lists wraps that in `{ logic: "and", predicates: [, { operator: "list_nin", expertListId }…] }`. Other combinations (intersections, nested exclusions) are rejected. The platform stores the condition in canonical form (groups flattened, predicates deduplicated and sorted). Not available to teams on the older team ban list. `null` detaches the lists. Omitted when no lists are attached. humanQaTimeForTask: description: >- Expected human review time per task in seconds; absent when not specified. exclusiveMinimum: 0 type: number llmQaModelId: default: gpt-5.6-terra description: Model identifier for LLM quality checks. enum: - gpt-5.6-terra - gpt-5.6-luna - gpt-5.4 - gpt-5.4-mini - gpt-5 - gpt-5-mini - claude-sonnet-4-6 type: string qaMode: const: hybrid description: >- Review mode: `llm_only` uses LLM checks, `human_only` uses human review, and `hybrid` combines them. type: string smartSamplingPercent: default: 30 description: >- Percentage of answers sampled for human review, from 10 to 100; defaults to 100 for human-only review and 30 for hybrid review. maximum: 100 minimum: 10 type: number required: - qaMode type: object - additionalProperties: false description: How answers are checked for quality. properties: qaMode: const: customer_review description: >- Review mode: `customer_review` has your team accept or reject every task of the node; an LLM QA check verifies the evidence of each reject. Only for teams with customer review enabled. type: string required: - qaMode type: object qualityCriteria: $ref: '#/components/schemas/QualityCriteria' reuseAudienceWithNodeId: anyOf: - additionalProperties: false description: >- Node whose expert audience is reused, or null when no audience is reused. properties: nodeId: description: Node ID whose qualified experts are reused. type: string pipelineId: description: >- Pipeline ID with prefix `pl.` containing the source audience. pattern: >- ^pl\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string required: - pipelineId - nodeId type: object - description: >- Node whose expert audience is reused, or null when no audience is reused. type: 'null' description: >- Node whose expert audience is reused, or null when no audience is reused. smartSearch: $ref: '#/components/schemas/SmartSearch' ui: $ref: '#/components/schemas/LabelingInterface' required: - about - publicName - entities - qualityCriteria - instruction - ui - onboarding type: object MatchNodeSettings: additionalProperties: false description: Configuration for this node type. properties: about: $ref: '#/components/schemas/NodeAbout' conditions: description: Routing conditions evaluated in order. items: additionalProperties: false description: One routing condition. properties: logic: default: and description: >- How predicates combine: `and` requires all to match; `or` requires at least one; defaults to `and`. enum: - and - or type: string name: description: Display name of the condition. type: string predicates: description: Predicates used by this routing condition. items: additionalProperties: false description: Condition applied to a field value. properties: fieldProxyId: description: Identifier of the field whose value is compared. type: string jsonPath: description: JSONPath locating the value within the field. type: string operator: description: >- Text comparison: `text_eq` for equality or `text_neq` for inequality. enum: - text_eq - text_neq type: string value: description: >- Scalar value compared as text; null is treated as an empty string. type: - string - number - boolean - 'null' required: - fieldProxyId - jsonPath - operator - value type: object type: array required: - name - predicates type: object type: array fieldProxies: description: Input and output field bindings for the node. properties: input: description: Fields available to this node as input. items: additionalProperties: false description: Field binding. properties: id: description: Field identifier from `graph.fields`. type: string optional: description: Whether this field may be absent when the node executes. type: boolean required: - id - optional type: object type: array required: - input type: object required: - about - conditions - fieldProxies type: object QuorumNodeSettings: additionalProperties: false description: Configuration for this node type. properties: about: $ref: '#/components/schemas/NodeAbout' audienceAndPricing: $ref: '#/components/schemas/HumanAudienceAndPricing' checks: additionalProperties: $ref: '#/components/schemas/CheckTestCases' description: >- Configured test inputs grouped by check type; absent when no tests are configured. propertyNames: enum: - ui - llmqa - annotation_qa - qc_lint - ui_types_lint - exam_dataset type: string type: object entities: $ref: '#/components/schemas/TaskEntities' exam: $ref: '#/components/schemas/QualificationExam' instruction: description: Task instructions shown to experts. type: string offboardingRules: default: [] description: >- Rules for excluding experts based on answer quality; defaults to an empty array. items: additionalProperties: false description: One expert-exclusion rule. properties: agreementScale: description: >- Field used to measure agreement; absent when no field is selected. type: string id: description: Identifier of the rule, formatted as a UUID. format: uuid pattern: >- ^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$ type: string type: description: >- Rule type: `mv_agreement` checks agreement with the majority answer; absent for an unspecified rule. enum: - mv_agreement type: string required: - id type: object type: array onboarding: $ref: '#/components/schemas/Onboarding' overlap: description: Configured number of independent expert answers per item. maximum: 100 minimum: 1 type: integer publicName: description: Task name shown to experts. type: string qualityAssurance: default: llmQaModelId: gpt-5.6-terra qaMode: llm_only description: How answers are checked for quality. oneOf: - description: How answers are checked for quality. properties: llmQaModelId: default: gpt-5.6-terra description: Model identifier for LLM quality checks. enum: - gpt-5.6-terra - gpt-5.6-luna - gpt-5.4 - gpt-5.4-mini - gpt-5 - gpt-5-mini - claude-sonnet-4-6 type: string qaMode: const: llm_only description: >- Review mode: `llm_only` uses LLM checks, `human_only` uses human review, and `hybrid` combines them. type: string required: - qaMode type: object - description: How answers are checked for quality. properties: qaMode: const: none description: >- Review mode: `llm_only` uses LLM checks, `human_only` uses human review, and `hybrid` combines them. type: string required: - qaMode type: object qualityCriteria: $ref: '#/components/schemas/QualityCriteria' reuseAudienceWithNodeId: anyOf: - additionalProperties: false description: >- Node whose expert audience is reused, or null when no audience is reused. properties: nodeId: description: Node ID whose qualified experts are reused. type: string pipelineId: description: >- Pipeline ID with prefix `pl.` containing the source audience. pattern: >- ^pl\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string required: - pipelineId - nodeId type: object - description: >- Node whose expert audience is reused, or null when no audience is reused. type: 'null' description: >- Node whose expert audience is reused, or null when no audience is reused. smartSearch: $ref: '#/components/schemas/SmartSearch' ui: $ref: '#/components/schemas/LabelingInterface' required: - about - publicName - entities - qualityCriteria - instruction - ui - overlap - onboarding type: object DatasetSplitPartExclusions: description: Split-part selections to exclude; omit when no parts should be excluded. items: additionalProperties: false description: Split field and selected part names. properties: fieldId: description: Split field ID with prefix `df.`. pattern: >- ^df\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string names: description: Names of the split parts to select. items: maxLength: 64 pattern: ^[A-Za-z0-9_]+(?:[.-][A-Za-z0-9_]+)*$ type: string maxItems: 10 minItems: 1 type: array required: - fieldId - names type: object maxItems: 500 type: array DatasetFieldSourceConditions: description: >- Field-author conditions that selected rows must satisfy; defaults to no conditions. items: additionalProperties: false description: Required authors for a field. properties: fieldId: description: Dataset field ID with prefix `df.`. pattern: >- ^df\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string sources: description: Revision authors to match for this field. items: $ref: '#/components/schemas/FieldRevisionSource' minItems: 1 type: array required: - fieldId - sources type: object type: array DatasetFieldValueConditions: description: >- Field-value conditions that selected rows must satisfy; defaults to no conditions. items: additionalProperties: false description: Condition applied to a field value. properties: fieldId: description: Dataset field ID with prefix `df.` to compare. pattern: >- ^df\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string jsonPath: description: JSONPath locating the value within the field. minLength: 1 type: string latestOnly: default: true description: >- Whether to test only the latest revision from the allowed authors; false tests any revision, and the default is true. type: boolean operator: description: >- Text comparison: `text_eq` for equality or `text_neq` for inequality. enum: - text_eq - text_neq type: string sources: default: [] description: Allowed revision authors; an empty array allows any author. items: $ref: '#/components/schemas/FieldRevisionSource' type: array value: description: Scalar value compared as text; null is treated as an empty string. type: - string - number - boolean - 'null' required: - jsonPath - operator - value - fieldId type: object type: array DatasetSplitPartInclusion: additionalProperties: false description: Split parts to include; omit to include every part. properties: fieldId: description: Split field ID with prefix `df.`. pattern: >- ^df\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string names: description: Names of the split parts to select. items: maxLength: 64 pattern: ^[A-Za-z0-9_]+(?:[.-][A-Za-z0-9_]+)*$ type: string maxItems: 10 minItems: 1 type: array required: - fieldId - names type: object DatasetSliceSelection: additionalProperties: false description: >- Slice whose frozen rows and fields to read; omit to use live dataset rows. properties: sliceId: description: Dataset slice ID with prefix `dsl.`. pattern: >- ^dsl\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string required: - sliceId type: object NodeAbout: additionalProperties: false description: Node name and task context. properties: description: description: Optional explanation of the node's purpose. type: string name: description: Display name of the node. type: string projectType: description: >- Task type: `general` for labeling existing data or `data_collection` for collecting new data; absent when unspecified. enum: - general - data_collection type: string required: - name type: object CheckTestCases: description: Test cases for this check type. items: description: One configured check case. properties: id: description: Identifier of the check case. type: string input: anyOf: - description: >- Check input as JSON text, or JSON values encoded as strings and keyed by field ID. type: string - additionalProperties: type: string description: >- Check input as JSON text, or JSON values encoded as strings and keyed by field ID. propertyNames: type: string type: object description: >- Check input as JSON text, or JSON values encoded as strings and keyed by field ID. name: description: Display name of the check case. type: string required: - id - name - input type: object type: array HumanAudienceAndPricing: additionalProperties: false description: Who completes tasks and how attempts are priced. properties: audience: additionalProperties: false description: Expert eligibility filters. properties: countries: additionalProperties: false description: Eligible expert countries. properties: type: description: >- Matching rule; `or` accepts experts from any selected country. enum: - or type: string values: description: >- Country identifiers to include; an empty array applies no country restriction. items: type: string type: array required: - type - values type: object languages: additionalProperties: false description: Required expert languages. properties: type: description: >- Matching rule: `and` requires all selected languages; `or` requires at least one. enum: - or - and type: string values: description: >- Languages to match; an empty array applies no language restriction. items: description: Selected expert language. properties: value: description: Language identifier used by the audience filter. type: string required: - value type: object type: array required: - type - values type: object specializations: additionalProperties: false description: Required expert specializations. properties: type: description: >- Matching rule: `and` requires all selected specializations; `or` requires at least one. enum: - or - and type: string values: description: >- Specializations to match; an empty array applies no specialization restriction. items: additionalProperties: false description: Selected expert specialization. properties: qualificationId: description: Qualification ID returned with the specialization. type: string value: description: >- Specialization name returned by the specializations endpoint. type: string required: - value - qualificationId type: object type: array required: - type - values type: object required: - languages - specializations - countries type: object audienceType: const: human description: >- Task performer: `human` for experts or `synthetic` for model-generated answers. type: string expertListCondition: anyOf: - $ref: '#/components/schemas/ExpertListCondition' - type: 'null' description: >- The expert lists the node takes its experts from: experts in any of the `list_in` lists, but not in any of the `list_nin` lists. One list is `{ operator: "list_in", expertListId }` with an `exl.` id; several are `{ logic: "or", predicates: [list_in…] }`; excluding lists wraps that in `{ logic: "and", predicates: [, { operator: "list_nin", expertListId }…] }`. Other combinations (intersections, nested exclusions) are rejected. The platform stores the condition in canonical form (groups flattened, predicates deduplicated and sorted). Not available to teams on the older team ban list. `null` detaches the lists. Omitted when no lists are attached. perExpertCapEnabled: default: false description: >- Whether to limit how many tasks each expert may submit; defaults to false. type: boolean perExpertCapValue: default: 1 description: >- Maximum tasks per expert when the cap is enabled, from 1 to 50; defaults to 1. exclusiveMinimum: 0 maximum: 50 type: integer pricePerTask: description: Price per expert task in US dollars. type: number timeForTask: description: Expected task completion time in seconds. type: number verifiedExperts: description: Whether experts must have verified identities. type: boolean required: - verifiedExperts - audience - pricePerTask - timeForTask - audienceType type: object TaskEntities: additionalProperties: false description: Task input and output entities. properties: input: description: Entities supplied as task input. items: $ref: '#/components/schemas/TaskEntity' type: array output: description: Entities the expert or model must produce. items: $ref: '#/components/schemas/TaskEntity' type: array required: - input - output type: object QualificationExam: additionalProperties: false description: >- Qualification exam settings; omitted settings use a disabled one-task exam. properties: enabled: description: Whether experts must pass an exam before receiving regular tasks. type: boolean goldenDatasetId: description: >- Dataset ID with prefix `ds.` containing reference answers; absent when no reference dataset is selected. pattern: >- ^ds\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string gradedEntities: description: Entities and grading rules used in the exam. items: additionalProperties: false description: Grading rule for one entity. properties: customInstruction: description: Instruction for custom reference-answer comparison. type: string disabledCriterionIds: description: UUIDs of quality criteria excluded from exam grading. items: format: uuid pattern: >- ^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$ type: string type: array groundTruthType: description: >- Reference comparison: `direct` compares values directly; `custom` uses the supplied instruction. enum: - direct - custom type: string mode: description: >- Grading mode: `ground_truth` compares with reference answers, `quality_criteria` applies quality requirements, and `not_graded` excludes the entity from grading. enum: - ground_truth - quality_criteria - not_graded type: string targetEntity: description: Field identifier of the entity being graded. type: string required: - targetEntity - mode - groundTruthType - customInstruction - disabledCriterionIds type: object type: array passThreshold: description: Minimum number of exam tasks an expert must pass. maximum: 9007199254740991 minimum: 0 type: integer tasksPerExpert: description: Number of exam tasks each expert receives. maximum: 9007199254740991 minimum: 1 type: integer required: - enabled - gradedEntities - tasksPerExpert - passThreshold type: object Onboarding: description: Preparation content experts complete before working on tasks. properties: enabled: description: Whether experts must complete onboarding. type: boolean info: description: Overview of the task for experts. properties: estimatedTime: description: Estimated completion time as display text. type: string goal: description: Goal of the task. type: string role: description: Role the expert takes in the task. type: string successCriteria: description: Requirements for completing the task successfully. type: string required: - estimatedTime - role - goal - successCriteria type: object steps: description: Onboarding steps in display order. items: description: One onboarding step. properties: blocks: description: >- Content blocks in this step; absent when the step has no blocks. items: description: One onboarding content block. oneOf: - description: One onboarding content block. properties: estimatedTime: description: Estimated completion time as display text. type: string goal: description: Goal of the task. type: string id: description: Identifier of the onboarding block. type: string role: description: Role the expert takes in the task. type: string successCriteria: description: Requirements for completing the task successfully. type: string type: const: info description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string required: - id - type - estimatedTime - role - goal - successCriteria type: object - description: One onboarding content block. properties: content: description: >- Markdown content for this block; absent when no content is supplied. items: description: Markdown content fragment. properties: type: const: markdown description: Content format; always `markdown`. type: string value: description: Markdown text to display. type: string required: - type - value type: object type: array id: description: Identifier of the onboarding block. type: string title: description: Optional heading for the block. type: string type: const: text description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string required: - id - type type: object - description: One onboarding content block. properties: content: description: >- Markdown content for this block; absent when no content is supplied. items: description: Markdown content fragment. properties: type: const: markdown description: Content format; always `markdown`. type: string value: description: Markdown text to display. type: string required: - type - value type: object type: array id: description: Identifier of the onboarding block. type: string questions: description: Questions in this quiz block. items: description: Question used to check an expert's understanding. properties: correctAnswer: anyOf: - description: >- Correct option value, or an array of correct values for a multiple-choice question. type: string - description: >- Correct option value, or an array of correct values for a multiple-choice question. items: type: string type: array description: >- Correct option value, or an array of correct values for a multiple-choice question. correctAnswerFeedback: description: >- Feedback shown for a correct answer; absent when none is configured. type: string hint: description: Optional hint shown with the question. type: string id: description: Identifier of the quiz question. type: string incorrectAnswerFeedback: description: >- Feedback shown for an incorrect answer; absent when none is configured. type: string options: description: Answer choices shown to the expert. items: description: One answer choice. properties: label: description: Text displayed for the answer choice. type: string value: description: Value identifying the answer choice. type: string required: - value - label type: object type: array question: description: Question text shown to the expert. type: string title: description: Optional title displayed above the question. type: string type: description: >- Question type: `single-choice` accepts one option; `multiple-choice` accepts several. enum: - single-choice - multiple-choice type: string required: - id - type - question - options - correctAnswer type: object type: array title: description: Optional heading for the block. type: string type: const: quiz description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string required: - id - type - questions type: object - description: One onboarding content block. properties: content: description: >- Markdown content for this block; absent when no content is supplied. items: description: Markdown content fragment. properties: type: const: markdown description: Content format; always `markdown`. type: string value: description: Markdown text to display. type: string required: - type - value type: object type: array editorConfig: description: Optional settings for the practice-answer editor. properties: language: description: >- Editor language identifier; absent when no language is selected. type: string placeholder: description: >- Placeholder text shown in an empty editor; absent when none is configured. type: string rows: description: >- Number of visible editor rows; absent to use the editor default. type: number type: object editorType: description: Optional practice-answer editor type. type: string id: description: Identifier of the onboarding block. type: string imageUrl: description: >- Optional image URL illustrating the practice exercise. type: string title: description: Optional heading for the block. type: string type: const: practice description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string validators: description: >- Practice-answer validation rules; absent when none are configured. items: description: One practice-answer validation rule. properties: criteria: description: Conditions the practice answer must meet. type: string description: description: Explanation of what this rule checks. type: string id: description: Identifier of this validation rule. type: string name: description: Display name of the validation rule. type: string required: - id - name - description - criteria type: object type: array required: - id - type type: object - description: One onboarding content block. properties: examples: description: Interactive examples in this block. items: description: Interactive example shown during onboarding. properties: checks: description: >- Checks that evaluate the expert's answer to this example. items: description: Check used to grade an onboarding example. oneOf: - description: >- Check used to grade an onboarding example. properties: golden: $ref: '#/components/schemas/JsonValue' description: >- Expected JSON value for an exact check; null is a valid expected value. hint: description: >- Hint explaining how to satisfy the check. type: string path: description: Path to the answer value to check. type: string type: const: exact description: >- Check type: `exact` compares an expected value, `code` runs a validation function, and `llm` assesses the answer against criteria. type: string required: - type - path - golden - hint type: object - description: >- Check used to grade an onboarding example. properties: code: description: Validation code used by a code check. type: string hint: description: >- Hint explaining how to satisfy the check. type: string type: const: code description: >- Check type: `exact` compares an expected value, `code` runs a validation function, and `llm` assesses the answer against criteria. type: string required: - type - code - hint type: object - description: >- Check used to grade an onboarding example. properties: criteria: description: >- Criteria used to assess the example answer. type: string hint: description: >- Hint explaining how to satisfy the check. type: string imageUrl: description: Optional image URL used by the check. type: string paths: description: Answer paths to assess together. items: type: string type: array reference: description: Optional reference answer for the check. type: string type: const: llm description: >- Check type: `exact` compares an expected value, `code` runs a validation function, and `llm` assesses the answer against criteria. type: string required: - type - paths - criteria type: object maxItems: 10 type: array explanation: description: Optional explanation of the example answer. type: string id: description: Identifier of the example. type: string input: additionalProperties: $ref: '#/components/schemas/JsonValue' description: Example input values keyed by field name. propertyNames: type: string type: object sourceItemId: description: >- Source dataset item ID, when the example was taken from a dataset; otherwise absent. type: string required: - id - input - checks type: object maxItems: 10 type: array id: description: Identifier of the onboarding block. type: string title: description: Optional heading for the block. type: string type: const: examples description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string required: - id - type - examples type: object - description: One onboarding content block. properties: consentLabel: description: Text of the consent option the expert must accept. type: string content: description: >- Markdown content for this block; absent when no content is supplied. items: description: Markdown content fragment. properties: type: const: markdown description: Content format; always `markdown`. type: string value: description: Markdown text to display. type: string required: - type - value type: object type: array id: description: Identifier of the onboarding block. type: string title: description: Optional heading for the block. type: string type: const: consent description: >- Block kind: `info`, `text`, `consent`, `quiz`, `practice`, or interactive `examples`. type: string required: - id - type - consentLabel type: object type: array id: description: Identifier of the onboarding step. type: string required: - id type: object type: array required: - enabled - steps - info type: object ExpertListCondition: anyOf: - $ref: '#/components/schemas/ExpertListPredicate' - $ref: '#/components/schemas/ExpertListConditionGroup' QualityCriteria: description: Requirements used to assess the generated answers. items: description: Quality requirement applied to a task entity. properties: id: description: Unique criterion identifier, formatted as a UUID. format: uuid pattern: >- ^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$ type: string llmQaInstruction: description: >- Additional instruction for LLM quality checks; absent when none is provided. maxLength: 5000 type: string qualityElement: description: Description of the quality requirement. type: string targetEntity: description: Field identifier of the entity assessed by this criterion. type: string thresholds: description: Conditions for passing or failing this criterion. properties: fail: description: What makes an answer fail this criterion. type: string pass: description: What an acceptable answer must satisfy. type: string required: - pass - fail type: object required: - id - qualityElement - thresholds - targetEntity type: object type: array SmartSearch: anyOf: - additionalProperties: false description: >- Expert recruitment criteria, or null when no recruitment brief is configured. properties: briefDescription: description: Short description of the expert role. type: string extraHardRequirements: description: Mandatory qualifications for candidate experts. items: type: string type: array idealCandidateProfile: description: Description of an ideal candidate. type: string projectDescription: description: Project context for recruitment, or null when none is supplied. type: - string - 'null' relevanceCriteria: description: Criteria used to judge whether a candidate fits the task. items: type: string type: array sourceFingerprint: description: Fingerprint of the source brief; absent when none is recorded. type: string typicalJobTitles: description: Job titles commonly held by suitable experts. items: type: string type: array whoIsNotAFit: description: Reasons a candidate would be unsuitable. items: type: string type: array required: - briefDescription - projectDescription - extraHardRequirements - relevanceCriteria - typicalJobTitles - idealCandidateProfile - whoIsNotAFit type: object - description: >- Expert recruitment criteria, or null when no recruitment brief is configured. type: 'null' default: null description: >- Expert recruitment criteria, or null when no recruitment brief is configured. LabelingInterface: description: Labeling interface and its data transformations. properties: code: description: Source code for the labeling interface. type: string constants: description: >- JSON object encoded as a string, containing constants available to the interface. type: string data: description: JSON text containing sample data for the interface preview. type: string postprocess: default: return {}; description: >- JavaScript code that transforms submitted interface values; defaults to `return {};`. type: string preprocess: description: JavaScript code that prepares input data for the interface. type: string required: - code - data - preprocess - constants type: object FieldRevisionSource: description: Source that wrote a field revision. oneOf: - additionalProperties: false description: Source that wrote a field revision. properties: type: const: user description: >- Revision author: `user` for a user action or `system` for a platform operation. type: string userId: description: Identifier of the user who wrote the revision. minLength: 1 type: string required: - type - userId type: object - additionalProperties: false description: Source that wrote a field revision. properties: source: description: Identifier of the platform operation that wrote the revision. minLength: 1 type: string type: const: system description: >- Revision author: `user` for a user action or `system` for a platform operation. type: string required: - type - source type: object TaskEntity: description: Task entity bound to a pipeline field. oneOf: - additionalProperties: false description: Task entity bound to a pipeline field. properties: allowedLabels: description: >- Allowed classification labels; absent for entities without a fixed label set. items: additionalProperties: false description: One allowed classification label. properties: label: description: Label displayed to the expert. type: string value: description: Value stored for this label in the annotation. type: string required: - label - value type: object type: array description: description: Explanation of the entity and how it is used in the task. type: string dynamic: const: false description: Whether one item can contain multiple instances of this entity. type: boolean entityType: description: >- Semantic entity type, such as `PROMPT`, `RESPONSE`, `SINGLE_LABEL`, or `VIDEO`. type: string fieldProxyId: description: Identifier of the field containing this entity. type: string jsonPath: description: JSONPath locating entity values within the field. type: string maxDurationSeconds: description: >- Maximum expected media duration in seconds; absent when not applicable. exclusiveMinimum: 0 type: number maxEntities: description: >- Maximum expected instances of a dynamic entity per item; null or absent when no maximum is set. type: 'null' minDurationSeconds: description: >- Minimum expected media duration in seconds; absent when no minimum is set. exclusiveMinimum: 0 type: number optional: description: Whether the entity value may be omitted. type: boolean relations: description: >- Relations to other entities; omit when the entity has no relations. items: anyOf: - description: Relation to another entity. properties: id: description: Field identifier of the related entity. type: string rel_type: description: >- Relation: `context` supplies context, `references` cites another entity, `depends_on` requires it, `evaluates` assesses it, or `conditional` applies only when a condition matches. enum: - context - references - depends_on - evaluates type: string required: - id - rel_type type: object - description: Relation to another entity. properties: condition: description: >- Condition controlling whether the related entity applies. properties: operator: description: >- Comparison: `EQ` or `NEQ` for equality, `IN` or `NOT_IN` for list membership, or `NOT_EMPTY` for a nonempty value. enum: - EQ - NEQ - IN - NOT_IN - NOT_EMPTY type: string value: anyOf: - description: >- Value or values to compare with the related entity. type: string - description: >- Value or values to compare with the related entity. items: type: string type: array description: >- Value or values to compare with the related entity. required: - operator - value type: object id: description: Field identifier of the related entity. type: string rel_type: const: conditional description: >- Relation: `context` supplies context, `references` cites another entity, `depends_on` requires it, `evaluates` assesses it, or `conditional` applies only when a condition matches. type: string required: - id - rel_type - condition type: object description: Relation to another entity. type: array required: - fieldProxyId - entityType - optional - jsonPath - description - dynamic type: object - additionalProperties: false description: Task entity bound to a pipeline field. properties: allowedLabels: description: >- Allowed classification labels; absent for entities without a fixed label set. items: additionalProperties: false description: One allowed classification label. properties: label: description: Label displayed to the expert. type: string value: description: Value stored for this label in the annotation. type: string required: - label - value type: object type: array description: description: Explanation of the entity and how it is used in the task. type: string dynamic: const: true description: Whether one item can contain multiple instances of this entity. type: boolean entityType: description: >- Semantic entity type, such as `PROMPT`, `RESPONSE`, `SINGLE_LABEL`, or `VIDEO`. type: string fieldProxyId: description: Identifier of the field containing this entity. type: string jsonPath: description: JSONPath locating entity values within the field. type: string maxDurationSeconds: description: >- Maximum expected media duration in seconds; absent when not applicable. exclusiveMinimum: 0 type: number maxEntities: description: >- Maximum expected instances of a dynamic entity per item; null or absent when no maximum is set. type: - number - 'null' minDurationSeconds: description: >- Minimum expected media duration in seconds; absent when no minimum is set. exclusiveMinimum: 0 type: number optional: description: Whether the entity value may be omitted. type: boolean relations: description: >- Relations to other entities; omit when the entity has no relations. items: anyOf: - description: Relation to another entity. properties: id: description: Field identifier of the related entity. type: string rel_type: description: >- Relation: `context` supplies context, `references` cites another entity, `depends_on` requires it, `evaluates` assesses it, or `conditional` applies only when a condition matches. enum: - context - references - depends_on - evaluates type: string required: - id - rel_type type: object - description: Relation to another entity. properties: condition: description: >- Condition controlling whether the related entity applies. properties: operator: description: >- Comparison: `EQ` or `NEQ` for equality, `IN` or `NOT_IN` for list membership, or `NOT_EMPTY` for a nonempty value. enum: - EQ - NEQ - IN - NOT_IN - NOT_EMPTY type: string value: anyOf: - description: >- Value or values to compare with the related entity. type: string - description: >- Value or values to compare with the related entity. items: type: string type: array description: >- Value or values to compare with the related entity. required: - operator - value type: object id: description: Field identifier of the related entity. type: string rel_type: const: conditional description: >- Relation: `context` supplies context, `references` cites another entity, `depends_on` requires it, `evaluates` assesses it, or `conditional` applies only when a condition matches. type: string required: - id - rel_type - condition type: object description: Relation to another entity. type: array required: - fieldProxyId - entityType - optional - jsonPath - description - dynamic type: object ExpertListPredicate: additionalProperties: false properties: expertListId: pattern: >- ^exl\.[0-9a-f]{8}-[0-9a-f]{4}-[7][0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$ type: string operator: enum: - list_in - list_nin type: string required: - operator - expertListId type: object ExpertListConditionGroup: additionalProperties: false properties: logic: enum: - and - or type: string predicates: items: $ref: '#/components/schemas/ExpertListCondition' minItems: 1 type: array required: - logic - predicates type: object headers: RateLimit-Limit: description: The maximum number of requests allowed in the current window. schema: type: integer RateLimit-Policy: description: The rate limit policy (e.g., "60;w=60"). schema: type: string RateLimit-Remaining: description: The number of requests remaining in the current window. schema: type: integer RateLimit-Reset: description: The number of seconds until the current window resets. schema: type: integer X-Request-Id: description: >- The request id: the X-Request-Id you sent when it is 1-128 characters of letters, digits, ".", "_" or "-", otherwise a generated UUID. Present on every response, including errors; quote it when reporting a problem. schema: type: string securitySchemes: apiKeyAuth: description: >- Enter `ApiKey ` in Authorization, including the `ApiKey ` prefix. In cURL and Python samples, `` means this full value. The TypeScript SDK adds the prefix for you. in: header name: Authorization type: apiKey scopedTokenAuth: description: >- Enter the project-scoped token as `pst_...` in X-Scoped-Token, without an `ApiKey ` prefix. Create a token with POST /projects/{projectId}/tokens. in: header name: X-Scoped-Token type: apiKey ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.