> ## Documentation Index > Fetch the complete documentation index at: https://docs.deck.co/llms.txt > Use this file to discover all available pages before exploring further. # List task run screenshots > Returns a paginated list of agent screenshots captured during a task run, oldest first. Each item includes a pre-signed download URL and the agent's reasoning at the time of capture. ## OpenAPI ````yaml /api-reference/v2.json get /task-runs/{run_id}/screenshots openapi: 3.1.1 info: title: Deck API version: 2.0.0 servers: - url: https://api.deck.co/v2 description: Production security: - BearerAuth: [] tags: - name: Agents description: >- Agents are the AI workers that execute tasks against sources. Create and manage the agents in your organization. - name: Components description: >- Embeddable UI components. Create short-lived session tokens for the Auth Component. - name: Platform description: Organization-wide platform settings, such as the static IP range. - name: Credentials description: >- Credentials store the login details an agent uses at a source. Values are tokenized at rest and never returned in plaintext. - name: Event Destinations description: >- Where Deck delivers events: webhooks, SQS, Kinesis, S3, and more. Manage destinations and inspect delivery attempts. - name: Events description: >- Immutable records of things that happened in your organization, such as a task run completing or a credential being verified. - name: Sessions description: >- Sessions are agent sandbox lifetimes. A session is created for each task run (or explicitly) and counts toward your concurrency limit. - name: Sources description: >- Sources are the websites and portals agents work against. Register, inspect, reset, and delete them. - name: Storage description: >- Files produced by task runs: downloaded documents, exports, and extraction results. - name: Task Runs description: >- A task run is one execution of a task. Inspect status and output, retrieve files and screenshots, cancel runs, and answer interaction requests. - name: Tasks description: >- Tasks define reusable units of work for an agent: a goal, an input schema, and an output schema. Run them against sources with credentials. - name: Test description: Connectivity check for your API key. - name: Triggers description: Triggers run tasks automatically on a schedule or in response to events. - name: Workflows - name: Workflow Runs paths: /task-runs/{run_id}/screenshots: get: tags: - Task Runs summary: List task run screenshots description: >- Returns a paginated list of agent screenshots captured during a task run, oldest first. Each item includes a pre-signed download URL and the agent's reasoning at the time of capture. operationId: listTaskRunScreenshots parameters: - name: run_id in: path required: true schema: type: string pattern: ^trun_ example: trun_AbC123xYz456 description: Unique identifier for the task run (prefixed with `trun_`). - name: limit in: query schema: pattern: ^-?(?:0|[1-9]\d*)$ type: - integer - string format: int32 minimum: 1 maximum: 100 default: 20 example: 20 description: Maximum number of items to return (1-100, default 20). - name: cursor in: query schema: type: string description: Cursor from a previous response for pagination. responses: '200': description: OK content: text/plain: schema: $ref: >- #/components/schemas/PaginatedResponseOfListScreenshotItemResponse application/json: schema: $ref: >- #/components/schemas/PaginatedResponseOfListScreenshotItemResponse text/json: schema: $ref: >- #/components/schemas/PaginatedResponseOfListScreenshotItemResponse '400': description: Bad request — validation error or malformed input. content: application/json: schema: $ref: '#/components/schemas/ValidationErrorResponse' '401': description: Unauthorized. '403': description: Forbidden. '404': description: Not found — the requested resource does not exist. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '409': description: >- Conflict — the request is valid but cannot be completed in the current state. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '422': description: >- Unprocessable content — the request body is understood but contains invalid values. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '429': description: >- Too many requests — the organization-wide API rate limit was exceeded (`rate_limit_exceeded`). Retry with exponential backoff. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' '500': description: Internal server error. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' components: schemas: PaginatedResponseOfListScreenshotItemResponse: type: object properties: data: type: array items: $ref: '#/components/schemas/ListScreenshotItemResponse' description: Array of screenshot items for the current page, oldest first. has_more: type: boolean description: '`true` if there are more results beyond this page.' next_cursor: type: - 'null' - string description: >- Opaque cursor string to pass as the `cursor` query parameter on the next request. `null` when there are no more pages. request_id: type: - 'null' - string description: Unique identifier for the API request. required: - data - has_more ValidationErrorResponse: required: - errors type: object properties: type: type: - 'null' - string description: RFC 9110 problem type URI. title: type: - 'null' - string description: Short summary of the validation failure. status: type: - 'null' - integer description: HTTP status code. format: int32 errors: type: object additionalProperties: type: array items: type: string description: Validation errors grouped by field name. traceId: type: - 'null' - string description: Request trace identifier. description: Validation error returned for malformed requests (HTTP 400). ErrorResponse: required: - errors - request_id type: object properties: errors: type: array items: $ref: '#/components/schemas/TaskRunError' description: One or more errors that describe what went wrong. request_id: type: string description: Unique identifier for this request, useful for tracing. description: Standard error response. ListScreenshotItemResponse: required: - id - object - file_name - file_type - file_size - created_at type: object properties: id: type: string description: Unique identifier for the storage item, prefixed with `stor_`. pattern: ^stor_ example: stor_AbC123xYz456 object: type: string description: Always `screenshot`. file_name: type: string description: Original file name. file_type: type: string description: MIME type of the screenshot (e.g. `image/png`, `image/jpeg`). file_size: pattern: ^-?(?:0|[1-9]\d*)$ type: - integer - string format: int64 description: File size in bytes. url: type: - 'null' - string description: >- Pre-signed URL to download the screenshot. Time-limited; re-fetch the item if it expires. reasoning: type: - 'null' - string description: Agent reasoning captured at the time of this screenshot, if any. created_at: type: string format: date-time description: When the screenshot was captured. example: '2026-08-24T12:00:00Z' description: A single agent screenshot captured during a task run. TaskRunError: required: - type - code - message type: object properties: type: type: string description: >- Error category (e.g. `source`, `auth`, `task`, `rate_limit`). Determines the class of failure. code: type: string description: Machine-readable error code. Use this for programmatic handling. message: type: string description: >- Human-readable explanation. Do not rely on this for logic — it may change. field: type: - 'null' - string description: The input field that caused the error, when applicable. description: Error details for a failed task run. Same structure as API-level errors. securitySchemes: BearerAuth: type: http description: Secret key (sk_live_...) scheme: bearer bearerFormat: JWT ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.