> ## Documentation Index
> Fetch the complete documentation index at: https://docs.labelbox.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Get evaluation overview

> Returns one evaluation dashboard page, read at one consistent point in time. Current and prior windows are adjacent and equal; gaps in history produce explicit null coverage fields rather than misleading zeroes.



## OpenAPI

````yaml /managed-agents/openapi.yaml get /managed-agents/v1/evaluations/overview
openapi: 3.1.0
info:
  title: Managed Agents API
  version: 0.13.0
  description: >-
    The Recursion Managed Agents REST API. Authenticate with a Recursion API key
    as a bearer token. An organization-scoped key acts in its own organization
    and needs nothing else; a tenant-scoped key must also send
    `x-organization-id` with an organization id or `default`. Field names follow
    each operation's published schema.
servers:
  - url: https://api.recursion.labelbox.com
security:
  - bearerAuth: []
paths:
  /managed-agents/v1/evaluations/overview:
    get:
      tags:
        - Evaluations
      summary: Get evaluation overview
      description: >-
        Returns one evaluation dashboard page, read at one consistent point in
        time. Current and prior windows are adjacent and equal; gaps in history
        produce explicit null coverage fields rather than misleading zeroes.
      operationId: managedAgentsGetEvaluationOverview
      parameters:
        - description: >-
            Length of the current window. The prior window is the adjacent
            interval of equal length.
          in: query
          name: range
          schema:
            default: 28d
            description: >-
              Length of the current window. The prior window is the adjacent
              interval of equal length.
            enum:
              - 7d
              - 28d
              - 90d
            type: string
        - description: >-
            Immutable evaluation-agent version whose rubric and verdicts define
            this overview.
          in: query
          name: evaluator_agent_version_id
          required: true
          schema:
            description: >-
              Immutable evaluation-agent version whose rubric and verdicts
              define this overview.
            format: uuid
            type: string
        - description: >-
            Optionally narrow page-level metrics, table rows, and retained
            history to one target agent.
          in: query
          name: target_agent_id
          schema:
            description: >-
              Optionally narrow page-level metrics, table rows, and retained
              history to one target agent.
            format: uuid
            type: string
        - description: >-
            Repeat to explicitly compare up to five target agents; absent values
            select the page target or the first five whole-filter agents.
          in: query
          name: series_target_agent_id
          schema:
            description: >-
              Repeat to explicitly compare up to five target agents; absent
              values select the page target or the first five whole-filter
              agents.
            items:
              $ref: '#/components/schemas/ManagedAgentsEvaluationOverviewUUID'
            maxItems: 5
            type: array
            uniqueItems: true
        - description: >-
            Repeat to add scoped previous target-agent versions; automatic
            latest versions plus these values may total at most twelve traces.
          in: query
          name: series_previous_target_agent_version_id
          schema:
            description: >-
              Repeat to add scoped previous target-agent versions; automatic
              latest versions plus these values may total at most twelve traces.
            items:
              $ref: '#/components/schemas/ManagedAgentsEvaluationOverviewUUID'
            maxItems: 11
            type: array
            uniqueItems: true
        - description: Verdict population plotted for each selected target-agent version.
          in: query
          name: series_metric
          schema:
            default: overall
            description: Verdict population plotted for each selected target-agent version.
            enum:
              - overall
              - criterion
            type: string
        - description: >-
            Rubric key plotted when series_metric is criterion; required exactly
            for that metric.
          in: query
          name: series_criterion_key
          schema:
            description: >-
              Rubric key plotted when series_metric is criterion; required
              exactly for that metric.
            type: string
        - description: >-
            Opt into bounded adaptive chart positions from exact timestamps
            through days. Omit for the legacy fixed daily grid.
          in: query
          name: series_resolution
          schema:
            description: >-
              Opt into bounded adaptive chart positions from exact timestamps
              through days. Omit for the legacy fixed daily grid.
            enum:
              - adaptive
            type: string
        - description: Maximum target-agent rows to return. Defaults to 25.
          in: query
          name: limit
          schema:
            default: 25
            description: Maximum target-agent rows to return. Defaults to 25.
            maximum: 100
            minimum: 1
            type: integer
        - description: >-
            Opaque signed continuation bound to the organization, frozen
            watermark, range, evaluator, singular table target, and effective
            limit. Comparison agents, versions, and metric may change without
            resetting the table page. Expires after one hour.
          in: query
          name: page_token
          schema:
            description: >-
              Opaque signed continuation bound to the organization, frozen
              watermark, range, evaluator, singular table target, and effective
              limit. Comparison agents, versions, and metric may change without
              resetting the table page. Expires after one hour.
            type: string
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsEvaluationOverviewResponse'
          description: >-
            Complete fixed-window evaluation Overview plus an optional signed
            target-agent continuation.
        '400':
          description: The request was rejected by schema or semantic validation.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiError'
        '401':
          description: The caller is unauthenticated.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiError'
        '403':
          description: The caller lacks a required permission or capability.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiErrorForbidden'
        '404':
          description: No such resource is reachable for this caller.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiErrorNotFound'
        '429':
          headers:
            Retry-After:
              description: Seconds the caller should wait before retrying.
              schema:
                type: string
              style: simple
          description: A rate limit is exhausted. Honor `Retry-After`.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiError'
        '500':
          description: An unexpected server-side failure.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiError'
        '502':
          description: The service returned an invalid response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiErrorBadGateway'
        '503':
          headers:
            Retry-After:
              description: Seconds the caller should wait before retrying.
              schema:
                type: string
              style: simple
          description: A dependency is unavailable. Retryable.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiError'
        '504':
          headers:
            Retry-After:
              description: When to retry, as delay seconds or an HTTP date.
              schema:
                type: string
          description: The service did not respond before the timeout. Retryable.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ManagedAgentsApiErrorGatewayTimeout'
components:
  schemas:
    ManagedAgentsEvaluationOverviewUUID:
      description: >-
        One canonical UUID identifying an organization-scoped comparison agent
        or immutable version.
      format: uuid
      type: string
      example: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
    ManagedAgentsEvaluationOverviewResponse:
      additionalProperties: false
      description: >-
        Complete fixed-window evaluation Overview plus an optional signed
        target-agent continuation.
      properties:
        agent_rows:
          description: Ranked target-agent metric rows for this signed cursor page.
          items:
            $ref: '#/components/schemas/ManagedAgentsAgentRow'
          type: array
        as_of:
          description: Durable analytics watermark freezing every query in this response.
          format: date-time
          type: string
        change_markers:
          description: >-
            Immutable agent-version and model markers when one target agent is
            selected.
          items:
            $ref: '#/components/schemas/ManagedAgentsChangeMarker'
          type: array
        change_markers_status:
          description: Whether markers are complete or require selecting one target agent.
          enum:
            - complete
            - select_target_agent
          type: string
        comparison:
          $ref: '#/components/schemas/ManagedAgentsEvaluationOverviewComparison'
          description: >-
            Bounded multi-agent and immutable-version traces for one selected
            metric.
        criterion_columns:
          description: >-
            Stable key order used to align every agent row's criterion_metrics
            array.
          items:
            type: string
          type: array
        criterion_series:
          description: >-
            Aligned criterion verdict series for the complete frozen window;
            legacy responses use a daily grid.
          items:
            $ref: '#/components/schemas/ManagedAgentsCriterionSeriesPoint'
          type: array
        criterion_series_granularity:
          description: >-
            Adaptive position semantics: exact timestamps are criterion-local,
            while interval granularities use shared aligned positions; omitted
            for the legacy daily response and empty adaptive collections.
          enum:
            - exact
            - second
            - minute
            - hour
            - day
          type: string
        from:
          description: Inclusive UTC start of the current Overview window.
          format: date-time
          type: string
        next_page_token:
          description: >-
            Signed one-hour continuation for the next ranked target-agent page
            at the same watermark.
          type: string
        prior_from:
          description: Inclusive UTC start of the adjacent equal-length prior window.
          format: date-time
          type: string
        prior_to:
          description: Exclusive UTC end of the prior window, equal to from.
          format: date-time
          type: string
        tiles:
          $ref: '#/components/schemas/ManagedAgentsEvaluationOverviewTiles'
          description: Fixed headline coverage, verdict, criterion, and exact-cost metrics.
        to:
          description: Exclusive UTC end of the current window, equal to as_of.
          format: date-time
          type: string
      required:
        - as_of
        - from
        - to
        - prior_from
        - prior_to
        - tiles
        - criterion_columns
        - agent_rows
        - criterion_series
        - change_markers
        - change_markers_status
        - comparison
      type: object
      example:
        agent_rows:
          - criterion_metrics:
              - null
            evaluation_count: 1
            newest_failures:
              - created_at: '2026-02-18T09:30:00.000Z'
                evaluation_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
                failed_criterion_keys:
                  - example
                snapshot_event_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
                target_session_available: true
                target_session_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            overall_pass:
              delta_pp: 1.5
              fail_count: 1
              not_applicable_count: 1
              pass_count: 1
              prior_rate: 1
              rate: 1
            target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        as_of: '2026-02-18T09:30:00.000Z'
        change_markers:
          - agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            at: '2026-02-18T09:30:00.000Z'
            from_model: example
            kind: agent_version
            target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            to_model: example
            version_number: 1
        change_markers_status: complete
        comparison:
          agents:
            - default_target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              display_name: example-name
              target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              version_options:
                - created_at: '2026-02-18T09:30:00.000Z'
                  has_current_observations: true
                  identity_status: catalog
                  is_latest: true
                  selected: true
                  target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
                  version_number: 1
              version_options_truncated: true
          criterion_key: example
          granularity: exact
          markers:
            - agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              at: '2026-02-18T09:30:00.000Z'
              from_model: example
              kind: agent_version
              target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              to_model: example
              version_number: 1
          metric: overall
          selection_source: default
          series:
            - created_at: '2026-02-18T09:30:00.000Z'
              identity_status: catalog
              is_latest: true
              points:
                - bucket_start: '2026-02-18T09:30:00.000Z'
                  fail_count: 1
                  not_applicable_count: 1
                  pass_count: 1
                  rate: 1
              target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
              version_number: 1
        criterion_columns:
          - example
        criterion_series:
          - bucket_start: '2026-02-18T09:30:00.000Z'
            criterion_key: example
            fail_count: 1
            not_applicable_count: 1
            pass_count: 1
            rate: 1
        criterion_series_granularity: exact
        from: '2026-02-18T09:30:00.000Z'
        next_page_token: example
        prior_from: '2026-02-18T09:30:00.000Z'
        prior_to: '2026-02-18T09:30:00.000Z'
        tiles:
          coverage:
            delta_pp: 1.5
            eligible_session_count: 1
            evaluated_eligible_session_count: 1
            evaluation_snapshot_count: 1
            history_status: complete
            prior_rate: 1
            rate: 1
          evaluation_cost:
            complete_count: 1
            completeness: complete
            delta_usd: example
            evaluation_count: 1
            per_evaluation_usd: example
            prior_per_evaluation_usd: example
            total_usd: example
          lowest_criterion: null
          overall_pass:
            delta_pp: 1.5
            fail_count: 1
            not_applicable_count: 1
            pass_count: 1
            prior_rate: 1
            rate: 1
        to: '2026-02-18T09:30:00.000Z'
    ManagedAgentsApiError:
      type: object
      properties:
        code:
          type: string
          minLength: 1
          description: Stable machine-readable error code.
          enum:
            - analytics
            - analytics_not_ready
            - artifact_fetch_unavailable
            - artifact_generation_mismatch
            - artifact_hash_mismatch
            - artifact_invalid_events
            - artifact_invalid_json
            - artifact_not_allowlisted
            - artifact_size_mismatch
            - atomic_ingest_timeout
            - auth_unavailable
            - automation_event_source_invalid
            - automation_memory_unavailable
            - automation_reference_invalid
            - automation_runs_unconfigured
            - automations_unconfigured
            - bad_gateway
            - built_in_catalog_unavailable
            - built_in_integrations_unconfigured
            - compute_offerings_unavailable
            - conflict
            - content_store_request_canceled
            - content_store_unavailable
            - content_store_upload_timeout
            - credential_encryption_unconfigured
            - delegate_preparation_failed
            - entity_family_not_permitted
            - environment_not_verified
            - evaluation_cancellation_pending
            - evaluation_deletion_pending
            - event_admission_unavailable
            - event_ingress_unavailable
            - event_source_verification_credential_invalid
            - event_sources_unconfigured
            - execution_unconfigured
            - file_quota_exceeded
            - file_storage_unconfigured
            - forbidden
            - gateway_timeout
            - gestalt_unconfigured
            - github_rate_limited
            - github_unavailable
            - handoff_access_invalid
            - handoff_access_unavailable
            - handoff_access_unsupported
            - handoff_capture_unavailable
            - handoff_requires_retest
            - handoff_revoke_unavailable
            - handoff_signal_unavailable
            - idempotency_conflict
            - idempotency_in_progress
            - idempotency_unavailable
            - ingest_policy_changed
            - ingest_policy_not_found
            - ingest_policy_revoked
            - integrations_unconfigured
            - internal_error
            - invalid_json
            - invalid_pull_request_review_result
            - invalid_request
            - invalid_scope
            - invariant_violation
            - live_fleet_unconfigured
            - managed_agents_unavailable
            - missing_dependency
            - model_gateway_metadata_missing
            - model_gateway_unconfigured
            - model_gateway_unreachable
            - mutation_transaction_budget_exceeded
            - not_found
            - page_token_expired
            - payload_too_large
            - persistence_busy
            - precondition_failed
            - precondition_required
            - provider_registration_unsupported
            - provider_selection_unsupported
            - rate_limit_exceeded
            - rate_limited
            - repository_automation_unconfigured
            - request_too_large
            - reserved_source_type
            - revision_conflict
            - run_log_store_unavailable
            - runner_image_resolution_failed
            - sandbox_provider_disabled
            - service_unavailable
            - session_analyst_busy
            - session_analyst_model_unavailable
            - session_analyst_unconfigured
            - session_start_admission_unavailable
            - session_start_not_admitted
            - setup_fallback_environment_unsafe
            - setup_run_finished
            - setup_run_in_progress
            - setup_run_limit
            - setup_run_not_started
            - setup_runs_unavailable
            - skill_storage_unconfigured
            - slack_channels_rejected
            - slack_channels_unavailable
            - slack_channels_unconfigured
            - slack_connection_malformed
            - slack_event_intake_unavailable
            - slack_installation_mismatch
            - slack_missing_scope
            - slack_rate_limited
            - slack_reapproval_required
            - source_ownership_mismatch
            - source_revision_conflict
            - spanner_unavailable
            - streaming_unsupported
            - submission_transaction_budget_exceeded
            - task_version_number_conflict
            - team_board_unconfigured
            - too_many_streams
            - unauthorized
            - unsupported_media_type
            - vault_creation_retired
            - version_content_conflict
            - webhook_delivery_too_large
            - workspace_boundary_denied
        message:
          type: string
          minLength: 1
          description: Human-readable error message.
        details:
          description: Optional structured error details.
          type: object
          propertyNames:
            type: string
          additionalProperties: {}
      required:
        - code
        - message
      additionalProperties: false
      description: Standard flat error response.
    ManagedAgentsApiErrorForbidden:
      type: object
      properties:
        code:
          type: string
          minLength: 1
          description: Stable machine-readable error code.
          enum:
            - forbidden
        message:
          type: string
          minLength: 1
          description: Human-readable error message.
        details:
          description: Optional structured error details.
          type: object
          propertyNames:
            type: string
          additionalProperties: {}
      required:
        - code
        - message
      additionalProperties: false
      description: Standard flat error response.
    ManagedAgentsApiErrorNotFound:
      type: object
      properties:
        code:
          type: string
          minLength: 1
          description: Stable machine-readable error code.
          enum:
            - not_found
        message:
          type: string
          minLength: 1
          description: Human-readable error message.
        details:
          description: Optional structured error details.
          type: object
          propertyNames:
            type: string
          additionalProperties: {}
      required:
        - code
        - message
      additionalProperties: false
      description: Standard flat error response.
    ManagedAgentsApiErrorBadGateway:
      type: object
      properties:
        code:
          type: string
          minLength: 1
          description: Stable machine-readable error code.
          enum:
            - bad_gateway
        message:
          type: string
          minLength: 1
          description: Human-readable error message.
        details:
          description: Optional structured error details.
          type: object
          propertyNames:
            type: string
          additionalProperties: {}
      required:
        - code
        - message
      additionalProperties: false
      description: Standard flat error response.
    ManagedAgentsApiErrorGatewayTimeout:
      type: object
      properties:
        code:
          type: string
          minLength: 1
          description: Stable machine-readable error code.
          enum:
            - gateway_timeout
        message:
          type: string
          minLength: 1
          description: Human-readable error message.
        details:
          description: Optional structured error details.
          type: object
          propertyNames:
            type: string
          additionalProperties: {}
      required:
        - code
        - message
      additionalProperties: false
      description: Standard flat error response.
    ManagedAgentsAgentRow:
      additionalProperties: false
      description: >-
        One target-agent row with overall and aligned criterion metrics plus
        newest failures.
      properties:
        criterion_metrics:
          description: >-
            Metrics aligned positionally with criterion_columns; null marks no
            observations for that key.
          items:
            anyOf:
              - type: 'null'
              - $ref: '#/components/schemas/ManagedAgentsVerdictMetric'
          type: array
        evaluation_count:
          description: >-
            Immutable evaluation snapshots for this target agent in the current
            window.
          format: int64
          minimum: 0
          type: integer
        newest_failures:
          description: Up to three newest failing evaluations for this target agent.
          items:
            $ref: '#/components/schemas/ManagedAgentsNewestFailure'
          type: array
        overall_pass:
          $ref: '#/components/schemas/ManagedAgentsVerdictMetric'
          description: Overall current and prior pass metric for this target agent.
        target_agent_id:
          description: Target agent represented by this Overview table row.
          format: uuid
          type: string
      required:
        - target_agent_id
        - evaluation_count
        - overall_pass
        - criterion_metrics
        - newest_failures
      type: object
      example:
        criterion_metrics:
          - null
        evaluation_count: 1
        newest_failures:
          - created_at: '2026-02-18T09:30:00.000Z'
            evaluation_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            failed_criterion_keys:
              - example
            snapshot_event_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            target_session_available: true
            target_session_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        overall_pass:
          delta_pp: 1.5
          fail_count: 1
          not_applicable_count: 1
          pass_count: 1
          prior_rate: 1
          rate: 1
        target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
    ManagedAgentsChangeMarker:
      additionalProperties: false
      description: >-
        Immutable target-agent version or model-change marker aligned with the
        Overview timeline.
      properties:
        agent_version_id:
          description: Immutable target-agent version beginning at this marker.
          format: uuid
          type: string
        at:
          description: UTC timestamp when this immutable agent version was created.
          format: date-time
          type: string
        from_model:
          description: >-
            Model used by the immediately preceding version for a model-change
            marker.
          type: string
        kind:
          description: >-
            Whether this marker denotes any version boundary or specifically a
            model change.
          enum:
            - agent_version
            - model_change
          type: string
        target_agent_id:
          description: Selected target agent whose immutable history supplies the marker.
          format: uuid
          type: string
        to_model:
          description: Model introduced by this version for a model-change marker.
          type: string
        version_number:
          description: Monotonic human-readable version number of the target agent.
          format: int64
          minimum: 1
          type: integer
      required:
        - kind
        - target_agent_id
        - agent_version_id
        - version_number
        - at
      type: object
      example:
        agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        at: '2026-02-18T09:30:00.000Z'
        from_model: example
        kind: agent_version
        target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        to_model: example
        version_number: 1
    ManagedAgentsEvaluationOverviewComparison:
      additionalProperties: false
      description: >-
        A bounded multi-agent, immutable-version evaluation comparison over one
        selected metric.
      properties:
        agents:
          description: >-
            Effective organization-scoped agents and their bounded version
            options.
          items:
            $ref: >-
              #/components/schemas/ManagedAgentsEvaluationOverviewComparisonAgent
          type: array
        criterion_key:
          description: Selected rubric key when metric is criterion; omitted for overall.
          type: string
        granularity:
          description: >-
            Adaptive position semantics: exact timestamps are trace-local, while
            interval granularities use shared aligned positions; omitted for the
            legacy daily response and empty adaptive collections.
          enum:
            - exact
            - second
            - minute
            - hour
            - day
          type: string
        markers:
          description: Version and model-change markers for the selected trace versions.
          items:
            $ref: '#/components/schemas/ManagedAgentsChangeMarker'
          type: array
        metric:
          description: >-
            Whether traces aggregate overall evaluation verdicts or one
            criterion.
          enum:
            - overall
            - criterion
          type: string
        selection_source:
          description: >-
            Whether the server chose the effective agents or the request named
            them explicitly.
          enum:
            - default
            - explicit
          type: string
        series:
          description: >-
            At most twelve selected agent-version traces in stable agent and
            version order.
          items:
            $ref: >-
              #/components/schemas/ManagedAgentsEvaluationOverviewComparisonSeries
          type: array
      required:
        - metric
        - selection_source
        - agents
        - series
        - markers
      type: object
      example:
        agents:
          - default_target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            display_name: example-name
            target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            version_options:
              - created_at: '2026-02-18T09:30:00.000Z'
                has_current_observations: true
                identity_status: catalog
                is_latest: true
                selected: true
                target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
                version_number: 1
            version_options_truncated: true
        criterion_key: example
        granularity: exact
        markers:
          - agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            at: '2026-02-18T09:30:00.000Z'
            from_model: example
            kind: agent_version
            target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            to_model: example
            version_number: 1
        metric: overall
        selection_source: default
        series:
          - created_at: '2026-02-18T09:30:00.000Z'
            identity_status: catalog
            is_latest: true
            points:
              - bucket_start: '2026-02-18T09:30:00.000Z'
                fail_count: 1
                not_applicable_count: 1
                pass_count: 1
                rate: 1
            target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            version_number: 1
    ManagedAgentsCriterionSeriesPoint:
      additionalProperties: false
      description: >-
        One aligned criterion-verdict series point in the frozen Overview
        window.
      properties:
        bucket_start:
          description: >-
            Exact evaluation time or inclusive UTC start of this aligned
            interval; legacy responses use window-anchored 24-hour buckets.
          format: date-time
          type: string
        criterion_key:
          description: Stable rubric key represented by this series point.
          type: string
        fail_count:
          description: Criterion fail verdicts at this exact time or in this interval.
          format: int64
          minimum: 0
          type: integer
        not_applicable_count:
          description: >-
            Criterion not-applicable verdicts at this exact time or in this
            interval.
          format: int64
          minimum: 0
          type: integer
        pass_count:
          description: Criterion pass verdicts at this exact time or in this interval.
          format: int64
          minimum: 0
          type: integer
        rate:
          description: Pass divided by pass plus fail, or null at a zero denominator.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
      required:
        - bucket_start
        - criterion_key
        - pass_count
        - fail_count
        - not_applicable_count
        - rate
      type: object
      example:
        bucket_start: '2026-02-18T09:30:00.000Z'
        criterion_key: example
        fail_count: 1
        not_applicable_count: 1
        pass_count: 1
        rate: 1
    ManagedAgentsEvaluationOverviewTiles:
      additionalProperties: false
      description: >-
        The four fixed-cardinality headline tiles for an evaluation Overview
        window.
      properties:
        coverage:
          $ref: '#/components/schemas/ManagedAgentsCoverageTile'
          description: Immutable-cohort coverage for the selected windows.
        evaluation_cost:
          $ref: '#/components/schemas/ManagedAgentsEvaluationCost'
          description: Exact current and prior per-evaluation cost comparison.
        lowest_criterion:
          anyOf:
            - type: 'null'
            - $ref: '#/components/schemas/ManagedAgentsLowestCriterion'
          description: >-
            Worst defined criterion metric, or null when none has a pass/fail
            denominator.
        overall_pass:
          $ref: '#/components/schemas/ManagedAgentsVerdictMetric'
          description: >-
            Overall evaluation verdict performance for the current and prior
            windows.
      required:
        - coverage
        - overall_pass
        - lowest_criterion
        - evaluation_cost
      type: object
      example:
        coverage:
          delta_pp: 1.5
          eligible_session_count: 1
          evaluated_eligible_session_count: 1
          evaluation_snapshot_count: 1
          history_status: complete
          prior_rate: 1
          rate: 1
        evaluation_cost:
          complete_count: 1
          completeness: complete
          delta_usd: example
          evaluation_count: 1
          per_evaluation_usd: example
          prior_per_evaluation_usd: example
          total_usd: example
        lowest_criterion: null
        overall_pass:
          delta_pp: 1.5
          fail_count: 1
          not_applicable_count: 1
          pass_count: 1
          prior_rate: 1
          rate: 1
    ManagedAgentsVerdictMetric:
      additionalProperties: false
      description: >-
        Pass, fail, not-applicable, rate, and adjacent-window comparison for one
        verdict population.
      properties:
        delta_pp:
          description: >-
            Current rate minus prior rate in percentage points, or null when
            either rate is null.
          maximum: 100
          minimum: -100
          type:
            - 'null'
            - number
        fail_count:
          description: Immutable fail verdicts in the current window.
          format: int64
          minimum: 0
          type: integer
        not_applicable_count:
          description: Immutable not-applicable verdicts in the current window.
          format: int64
          minimum: 0
          type: integer
        pass_count:
          description: Immutable pass verdicts in the current window.
          format: int64
          minimum: 0
          type: integer
        prior_rate:
          description: >-
            Equivalent pass rate for the adjacent prior window, or null at zero
            denominator.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
        rate:
          description: Pass divided by pass plus fail, or null at a zero denominator.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
      required:
        - pass_count
        - fail_count
        - not_applicable_count
        - rate
        - prior_rate
        - delta_pp
      type: object
      example:
        delta_pp: 1.5
        fail_count: 1
        not_applicable_count: 1
        pass_count: 1
        prior_rate: 1
        rate: 1
    ManagedAgentsNewestFailure:
      additionalProperties: false
      description: >-
        One of the newest failing evaluations for a target agent, with deep-link
        evidence identity.
      properties:
        created_at:
          description: UTC timestamp when the failing evaluation was recorded.
          format: date-time
          type: string
        evaluation_id:
          description: Immutable failing evaluation to open from the Overview.
          format: uuid
          type: string
        failed_criterion_keys:
          description: Key-sorted rubric criteria that failed in this evaluation.
          items:
            type: string
          type: array
        snapshot_event_id:
          description: >-
            Canonical UUID transcript boundary to select when opening evidence;
            may be a current UUIDv7 or a retained legacy UUIDv4 event id.
          format: uuid
          type: string
        target_session_available:
          description: >-
            Whether the same-organization target session is still available to
            open.
          type: boolean
        target_session_id:
          description: Target session linked from the failure.
          format: uuid
          type: string
      required:
        - evaluation_id
        - target_session_id
        - snapshot_event_id
        - target_session_available
        - created_at
        - failed_criterion_keys
      type: object
      example:
        created_at: '2026-02-18T09:30:00.000Z'
        evaluation_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        failed_criterion_keys:
          - example
        snapshot_event_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        target_session_available: true
        target_session_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
    ManagedAgentsEvaluationOverviewComparisonAgent:
      additionalProperties: false
      description: >-
        One selected target agent and its bounded immutable-version picker
        options.
      properties:
        default_target_agent_version_id:
          description: >-
            Newest immutable version before the frozen watermark, selected
            automatically.
          format: uuid
          type: string
        display_name:
          description: >-
            Stable human-readable name from the frozen latest catalog version,
            or Deleted agent when only retained history remains.
          type: string
        target_agent_id:
          description: >-
            Organization-scoped target agent represented by this comparison
            group.
          format: uuid
          type: string
        version_options:
          description: >-
            Latest plus a bounded preview of prior versions available for
            comparison.
          items:
            $ref: >-
              #/components/schemas/ManagedAgentsEvaluationOverviewComparisonVersion
          type: array
        version_options_truncated:
          description: >-
            Whether additional prior versions exist beyond this bounded option
            preview.
          type: boolean
      required:
        - target_agent_id
        - display_name
        - default_target_agent_version_id
        - version_options
        - version_options_truncated
      type: object
      example:
        default_target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        display_name: example-name
        target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        version_options:
          - created_at: '2026-02-18T09:30:00.000Z'
            has_current_observations: true
            identity_status: catalog
            is_latest: true
            selected: true
            target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
            version_number: 1
        version_options_truncated: true
    ManagedAgentsEvaluationOverviewComparisonSeries:
      additionalProperties: false
      description: One selected immutable target-version trace in an evaluation comparison.
      properties:
        created_at:
          description: >-
            Catalog version creation time, omitted for retained-history-only
            identities.
          format: date-time
          type: string
        identity_status:
          description: >-
            Whether catalog metadata or retained history proves this trace
            identity.
          enum:
            - catalog
            - retained_history
          type: string
        is_latest:
          description: Whether this was the newest version at the frozen watermark.
          type: boolean
        points:
          description: >-
            Exact timestamps observed for this trace, or shared aligned interval
            positions; legacy responses contain one fixed 24-hour point per day.
          items:
            $ref: >-
              #/components/schemas/ManagedAgentsEvaluationOverviewComparisonPoint
          type: array
        target_agent_id:
          description: Target agent owning this comparison trace.
          format: uuid
          type: string
        target_agent_version_id:
          description: >-
            Immutable target-agent version whose stored evaluations form this
            trace.
          format: uuid
          type: string
        version_number:
          description: >-
            Catalog version number, omitted for retained-history-only
            identities.
          format: int64
          minimum: 1
          type: integer
      required:
        - target_agent_id
        - target_agent_version_id
        - is_latest
        - identity_status
        - points
      type: object
      example:
        created_at: '2026-02-18T09:30:00.000Z'
        identity_status: catalog
        is_latest: true
        points:
          - bucket_start: '2026-02-18T09:30:00.000Z'
            fail_count: 1
            not_applicable_count: 1
            pass_count: 1
            rate: 1
        target_agent_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        version_number: 1
    ManagedAgentsCoverageTile:
      additionalProperties: false
      description: >-
        Evaluation coverage over the immutable eligible-session cohort, with
        explicit unavailable history.
      properties:
        delta_pp:
          description: >-
            Current coverage minus prior coverage in percentage points, or null
            when incomparable.
          maximum: 100
          minimum: -100
          type:
            - 'null'
            - number
        eligible_session_count:
          description: >-
            Immutable eligible target cohort size, or null when history is
            unavailable.
          minimum: 0
          type:
            - 'null'
            - integer
        evaluated_eligible_session_count:
          description: >-
            Eligible cohort sessions with at least one evaluation, or null when
            history is unavailable.
          minimum: 0
          type:
            - 'null'
            - integer
        evaluation_snapshot_count:
          description: Exact immutable evaluation rows in the current window.
          format: int64
          minimum: 0
          type: integer
        history_status:
          description: >-
            Whether immutable cohort history fully covers both requested
            windows.
          enum:
            - complete
            - unavailable
          type: string
        prior_rate:
          description: >-
            Coverage rate for the adjacent prior window, or null when
            unavailable or empty.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
        rate:
          description: >-
            Evaluated eligible sessions divided by eligible sessions, or null
            when unavailable or empty.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
      required:
        - evaluation_snapshot_count
        - eligible_session_count
        - evaluated_eligible_session_count
        - rate
        - prior_rate
        - delta_pp
        - history_status
      type: object
      example:
        delta_pp: 1.5
        eligible_session_count: 1
        evaluated_eligible_session_count: 1
        evaluation_snapshot_count: 1
        history_status: complete
        prior_rate: 1
        rate: 1
    ManagedAgentsEvaluationCost:
      additionalProperties: false
      description: >-
        Exact evaluation cost totals and per-evaluation comparison without
        floating-point USD loss.
      properties:
        complete_count:
          description: Evaluations whose complete priced cost contributes to the total.
          format: int64
          minimum: 0
          type: integer
        completeness:
          description: >-
            Aggregate cost-finality state, or null when the window has no
            evaluations.
          enum:
            - complete
            - pending
            - unpriced
            - partial
            - legacy_partial
            - indeterminate
            - null
          type:
            - 'null'
            - string
        delta_usd:
          description: >-
            Exact current minus prior per-evaluation USD, or null when either
            side is undefined.
          type:
            - 'null'
            - string
        evaluation_count:
          description: Immutable evaluations included in the current cost population.
          format: int64
          minimum: 0
          type: integer
        per_evaluation_usd:
          description: >-
            Exact current-window USD total per cost-complete evaluation, or null
            when undefined.
          type:
            - 'null'
            - string
        prior_per_evaluation_usd:
          description: >-
            Equivalent exact per-evaluation USD for the adjacent prior window,
            or null when undefined.
          type:
            - 'null'
            - string
        total_usd:
          description: >-
            Exact decimal USD total, or null when no complete priced total
            exists.
          type:
            - 'null'
            - string
      required:
        - total_usd
        - evaluation_count
        - complete_count
        - per_evaluation_usd
        - prior_per_evaluation_usd
        - delta_usd
        - completeness
      type: object
      example:
        complete_count: 1
        completeness: complete
        delta_usd: example
        evaluation_count: 1
        per_evaluation_usd: example
        prior_per_evaluation_usd: example
        total_usd: example
    ManagedAgentsLowestCriterion:
      additionalProperties: false
      description: >-
        The rubric criterion with the lowest defined pass rate in the current
        window.
      properties:
        criterion_key:
          description: Rubric key with the lowest defined current-window pass rate.
          type: string
        metric:
          $ref: '#/components/schemas/ManagedAgentsVerdictMetric'
          description: Current and prior verdict metric for the selected criterion.
      required:
        - criterion_key
        - metric
      type: object
      example:
        criterion_key: example
        metric:
          delta_pp: 1.5
          fail_count: 1
          not_applicable_count: 1
          pass_count: 1
          prior_rate: 1
          rate: 1
    ManagedAgentsEvaluationOverviewComparisonVersion:
      additionalProperties: false
      description: >-
        One bounded immutable target-version option for an evaluation
        comparison.
      properties:
        created_at:
          description: >-
            Catalog creation time, omitted when only retained history proves the
            identity.
          format: date-time
          type: string
        has_current_observations:
          description: >-
            Whether the selected evaluator observed this version in the current
            chart window.
          type: boolean
        identity_status:
          description: >-
            Whether live catalog metadata or retained same-organization history
            proves this identity.
          enum:
            - catalog
            - retained_history
          type: string
        is_latest:
          description: >-
            Whether this was the newest immutable version before the frozen
            watermark.
          type: boolean
        selected:
          description: >-
            Whether this version contributes a comparison trace in this
            response.
          type: boolean
        target_agent_version_id:
          description: Immutable target-agent version represented by this option.
          format: uuid
          type: string
        version_number:
          description: >-
            Catalog version number, omitted when only retained history proves
            the identity.
          format: int64
          minimum: 1
          type: integer
      required:
        - target_agent_version_id
        - is_latest
        - selected
        - has_current_observations
        - identity_status
      type: object
      example:
        created_at: '2026-02-18T09:30:00.000Z'
        has_current_observations: true
        identity_status: catalog
        is_latest: true
        selected: true
        target_agent_version_id: 9f8b1c2d-3e4f-5a6b-7c8d-9e0f1a2b3c4d
        version_number: 1
    ManagedAgentsEvaluationOverviewComparisonPoint:
      additionalProperties: false
      description: One aligned verdict aggregate for a selected target-agent version.
      properties:
        bucket_start:
          description: >-
            Exact evaluation time or inclusive UTC start of this aligned
            interval; legacy responses use window-anchored 24-hour buckets.
          format: date-time
          type: string
        fail_count:
          description: >-
            Fail verdicts for this version at this exact time or in this
            interval.
          format: int64
          minimum: 0
          type: integer
        not_applicable_count:
          description: >-
            Not-applicable verdicts for this version at this exact time or in
            this interval.
          format: int64
          minimum: 0
          type: integer
        pass_count:
          description: >-
            Pass verdicts for this version at this exact time or in this
            interval.
          format: int64
          minimum: 0
          type: integer
        rate:
          description: Pass divided by pass plus fail, or null at a zero denominator.
          maximum: 1
          minimum: 0
          type:
            - 'null'
            - number
      required:
        - bucket_start
        - pass_count
        - fail_count
        - not_applicable_count
        - rate
      type: object
      example:
        bucket_start: '2026-02-18T09:30:00.000Z'
        fail_count: 1
        not_applicable_count: 1
        pass_count: 1
        rate: 1
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: A Recursion API key, created in the console under API keys.

````