--- title: "Genai Spans Query" method: POST path: "/agents/spans/query" tags: ["Agents"] --- # Genai Spans Query `POST /agents/spans/query` ## Request - Base URL: `https://trace.wandb.ai` (the document declares no server; this is the origin it was published from) - URL: `https://trace.wandb.ai/agents/spans/query` - Auth: none declared Query agent spans, either as raw rows or grouped aggregates. ## Request body - AgentSpansQueryReq: Request to query agent spans for a project. When `group_by` is empty (or omitted), returns raw span rows in the response's `spans` field. When `group_by` is non-empty, returns aggregate group rows in the response's `groups` field. - `project_id` string, required - `query` Query - `$expr` union, required - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - `$and` union[], required - union - LiteralOperation: Represents a constant value in the query language. This can be any standard JSON-serializable value. Example: ``` {"$literal": "predict"} ``` - … - GetFieldOperator: Access a field on the traced call. Supports dot notation for nested access, e.g. `summary.usage.tokens`. Only works on fields present in the `CallSchema`, including: - Top-level fields like `op_name`, `trace_id`, `started_at` - Nested fields like `inputs.input_name`, `summary.usage.tokens`, etc. Example: ``` {"$getField": "op_name"} ``` - … - ConvertOperation: Convert the input value to a specific type (e.g., `int`, `bool`, `string`). Example: ``` { "$convert": { "input": {"$getField": "inputs.value"}, "to": "int" } } ``` - … - SizeOperation: Return the number of elements in a collection or characters in a string. Example: ``` {"$size": {"$getField": "scorer_tags"}} ``` - … - AndOperation (recursive) - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - … - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - … - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - … - ContainsOperation: Case-insensitive substring match. Not part of MongoDB. Weave-specific addition. Example: ``` { "$contains": { "input": {"$getField": "display_name"}, "substr": {"$literal": "llm"}, "case_insensitive": true } } ``` - … - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - `$or` union[], required - union - LiteralOperation: Represents a constant value in the query language. This can be any standard JSON-serializable value. Example: ``` {"$literal": "predict"} ``` - … - GetFieldOperator: Access a field on the traced call. Supports dot notation for nested access, e.g. `summary.usage.tokens`. Only works on fields present in the `CallSchema`, including: - Top-level fields like `op_name`, `trace_id`, `started_at` - Nested fields like `inputs.input_name`, `summary.usage.tokens`, etc. Example: ``` {"$getField": "op_name"} ``` - … - ConvertOperation: Convert the input value to a specific type (e.g., `int`, `bool`, `string`). Example: ``` { "$convert": { "input": {"$getField": "inputs.value"}, "to": "int" } } ``` - … - SizeOperation: Return the number of elements in a collection or characters in a string. Example: ``` {"$size": {"$getField": "scorer_tags"}} ``` - … - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - … - OrOperation (recursive) - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - … - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - … - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - … - ContainsOperation: Case-insensitive substring match. Not part of MongoDB. Weave-specific addition. Example: ``` { "$contains": { "input": {"$getField": "display_name"}, "substr": {"$literal": "llm"}, "case_insensitive": true } } ``` - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - `$not` unknown[], required - unknown - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - `$eq` unknown[], required - unknown - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gt` unknown[], required - unknown - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lt` unknown[], required - unknown - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gte` unknown[], required - unknown - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lte` unknown[], required - unknown - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - `$in` unknown[], required - unknown - ContainsOperation: Case-insensitive substring match. Not part of MongoDB. Weave-specific addition. Example: ``` { "$contains": { "input": {"$getField": "display_name"}, "substr": {"$literal": "llm"}, "case_insensitive": true } } ``` - `$contains` ContainsSpec, required: Specification for the `$contains` operation. - `input`: The string to search. - `substr`: The substring to search for. - `case_insensitive`: If true, match is case-insensitive. - `input` union, required - LiteralOperation: Represents a constant value in the query language. This can be any standard JSON-serializable value. Example: ``` {"$literal": "predict"} ``` - … - GetFieldOperator: Access a field on the traced call. Supports dot notation for nested access, e.g. `summary.usage.tokens`. Only works on fields present in the `CallSchema`, including: - Top-level fields like `op_name`, `trace_id`, `started_at` - Nested fields like `inputs.input_name`, `summary.usage.tokens`, etc. Example: ``` {"$getField": "op_name"} ``` - … - ConvertOperation: Convert the input value to a specific type (e.g., `int`, `bool`, `string`). Example: ``` { "$convert": { "input": {"$getField": "inputs.value"}, "to": "int" } } ``` - … - SizeOperation: Return the number of elements in a collection or characters in a string. Example: ``` {"$size": {"$getField": "scorer_tags"}} ``` - … - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - … - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - … - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - … - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - … - ContainsOperation (recursive) - `substr` union, required - LiteralOperation: Represents a constant value in the query language. This can be any standard JSON-serializable value. Example: ``` {"$literal": "predict"} ``` - … - GetFieldOperator: Access a field on the traced call. Supports dot notation for nested access, e.g. `summary.usage.tokens`. Only works on fields present in the `CallSchema`, including: - Top-level fields like `op_name`, `trace_id`, `started_at` - Nested fields like `inputs.input_name`, `summary.usage.tokens`, etc. Example: ``` {"$getField": "op_name"} ``` - … - ConvertOperation: Convert the input value to a specific type (e.g., `int`, `bool`, `string`). Example: ``` { "$convert": { "input": {"$getField": "inputs.value"}, "to": "int" } } ``` - … - SizeOperation: Return the number of elements in a collection or characters in a string. Example: ``` {"$size": {"$getField": "scorer_tags"}} ``` - … - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - … - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - … - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - … - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - … - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - … - ContainsOperation (recursive) - `case_insensitive` boolean, nullable - `group_by` AgentGroupByRef[], nullable - `source` 'field' | 'column' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `alias` string, nullable - `measures` AgentSpanMeasureSpec[] - `alias` string, required - `aggregation` 'sum' | 'avg' | 'min' | 'max' | 'count' | 'count_distinct' | 'count_true' | 'count_false', required - `value` AgentSpanValueRef: Reference to a span field or typed custom attribute map value. - `source` 'field' | 'derived' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `value_type` 'datetime' | 'number' | 'boolean' | 'string', nullable - `filter` Query - `$expr` union, required - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - `$and` union[], required - union - … - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - `$or` union[], required - union - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - `$not` unknown[], required - unknown - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - `$eq` unknown[], required - unknown - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gt` unknown[], required - unknown - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lt` unknown[], required - unknown - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gte` unknown[], required - unknown - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lte` unknown[], required - unknown - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - `$in` unknown[], required - unknown - ContainsOperation: Case-insensitive substring match. Not part of MongoDB. Weave-specific addition. Example: ``` { "$contains": { "input": {"$getField": "display_name"}, "substr": {"$literal": "llm"}, "case_insensitive": true } } ``` - `$contains` ContainsSpec, required: Specification for the `$contains` operation. - `input`: The string to search. - `substr`: The substring to search for. - `case_insensitive`: If true, match is case-insensitive. - `input` union, required - … - `substr` union, required - … - `case_insensitive` boolean, nullable - `group_filters` AgentSpanGroupFilter[] - `group_by` AgentGroupByRef[] - `source` 'field' | 'column' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `alias` string, nullable - `measure` AgentSpanMeasureSpec, required: One aggregate measure computed over spans in a group or bucket. - `alias` string, required - `aggregation` 'sum' | 'avg' | 'min' | 'max' | 'count' | 'count_distinct' | 'count_true' | 'count_false', required - `value` AgentSpanValueRef: Reference to a span field or typed custom attribute map value. - `source` 'field' | 'derived' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `value_type` 'datetime' | 'number' | 'boolean' | 'string', nullable - `filter` Query - `$expr` union, required - AndOperation: Logical AND. All conditions must evaluate to true. Example: ``` { "$and": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "predict"}]}, {"$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 1000}]} ] } ``` - `$and` union[], required - … - OrOperation: Logical OR. At least one condition must be true. Example: ``` { "$or": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "a"}]}, {"$eq": [{"$getField": "op_name"}, {"$literal": "b"}]} ] } ``` - `$or` union[], required - … - NotOperation: Logical NOT. Inverts the condition. Example: ``` { "$not": [ {"$eq": [{"$getField": "op_name"}, {"$literal": "debug"}]} ] } ``` - `$not` unknown[], required - … - EqOperation: Equality check between two operands. Example: ``` { "$eq": [{"$getField": "op_name"}, {"$literal": "predict"}] } ``` - `$eq` unknown[], required - … - GtOperation: Greater than comparison. Example: ``` { "$gt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gt` unknown[], required - … - LtOperation: Less than comparison. Example: ``` { "$lt": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lt` unknown[], required - … - GteOperation: Greater than or equal comparison. Example: ``` { "$gte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$gte` unknown[], required - … - LteOperation: Less than or equal comparison. Example: ``` { "$lte": [{"$getField": "summary.usage.tokens"}, {"$literal": 100}] } ``` - `$lte` unknown[], required - … - InOperation: Membership check. Returns true if the left operand is in the list provided as the second operand. Example: ``` { "$in": [ {"$getField": "op_name"}, [{"$literal": "predict"}, {"$literal": "generate"}] ] } ``` - `$in` unknown[], required - … - ContainsOperation: Case-insensitive substring match. Not part of MongoDB. Weave-specific addition. Example: ``` { "$contains": { "input": {"$getField": "display_name"}, "substr": {"$literal": "llm"}, "case_insensitive": true } } ``` - `$contains` ContainsSpec, required: Specification for the `$contains` operation. - `input`: The string to search. - `substr`: The substring to search for. - `case_insensitive`: If true, match is case-insensitive. - … - `min` union - number - string, date-time - `max` union - number - string, date-time - `group_distributions` AgentSpanGroupDistributionSpec[] - `alias` string, required - `value` AgentSpanValueRef, required: Reference to a span field or typed custom attribute map value. - `source` 'field' | 'derived' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `bins` integer - `top_n` integer - `custom_attr_columns` AgentSpanValueRef[] - `source` 'field' | 'derived' | 'custom_attrs_string' | 'custom_attrs_int' | 'custom_attrs_float' | 'custom_attrs_bool' - `key` string, required - `include_details` boolean - `include_costs` boolean - `sort_by` AgentSortBy[], nullable - `field` string, required - `direction` 'asc' | 'desc' - `limit` integer - `offset` integer - `started_after` string, date-time, nullable - `started_before` string, date-time, nullable - `signal_filters` AgentSignalFilter - `tags` string[] - `ratings` RatingCondition[] - `scorer_key` string, required - `op` 'gte' | 'gt' | 'lte' | 'lt' | 'eq', required - `value` number, required - `insight_filters` AgentInsightFilter[] - `field` 'intent_category' | 'intent_sentiment' | 'failure_category' | 'failure_severity' | 'intent_topic_id' | 'failure_topic_id', required - `values` string[], required: Values to match. intent_sentiment accepts only frustrated, dissatisfied, neutral, satisfied, delighted; failure_severity accepts only info, major, or minor. - `exclude` boolean: Exclude conversations matching any value in this filter. ## Response `200` Successful Response - AgentSpansQueryRes: Response from a spans query. Exactly one of `spans` or `groups` will be populated, based on whether the request specified `group_by`. - `spans` AgentSpanSchema[], required - `project_id` string, required - `trace_id` string, required - `span_id` string, required - `parent_span_id` string, nullable, required - `span_name` string, nullable, required - `span_kind` 'UNSPECIFIED' | 'INTERNAL' | 'SERVER' | 'CLIENT' | 'PRODUCER' | 'CONSUMER', nullable, required - `started_at` string, date-time, nullable, required - `ended_at` string, date-time, nullable, required - `status_code` 'UNSET' | 'OK' | 'ERROR', nullable, required - `status_message` string, nullable, required - `operation_name` string, nullable, required - `provider_name` string, nullable, required - `agent_name` string, nullable, required - `agent_id` string, nullable, required - `agent_description` string, nullable, required - `agent_version` string, nullable, required - `main_agent_id` string, nullable, required - `main_agent_name` string, nullable, required - `workflow_name` string, nullable, required - `eval_run_id` string, nullable, required - `eval_predict_and_score_call_id` string, nullable, required - `eval_kind` string, nullable, required - `eval_row_digest` string, nullable, required - `eval_example_id` string, nullable, required - `eval_trial_index` integer, nullable, required - `eval_evaluation_name` string, nullable, required - `parent_call_id` string, nullable, required - `parent_call_trace_id` string, nullable, required - `request_model` string, nullable, required - `response_model` string, nullable, required - `response_id` string, nullable, required - `response_time_to_first_chunk` number, nullable, required - `input_tokens` integer, nullable, required - `output_tokens` integer, nullable, required - `reasoning_tokens` integer, nullable, required - `cache_creation_input_tokens` integer, nullable, required - `cache_read_input_tokens` integer, nullable, required - `input_cost_usd` number, nullable, required - `output_cost_usd` number, nullable, required - `cache_read_cost_usd` number, nullable, required - `cache_creation_cost_usd` number, nullable, required - `total_cost_usd` number, nullable, required - `reasoning_content` string, nullable, required - `conversation_id` string, nullable, required - `conversation_name` string, nullable, required - `tool_name` string, nullable, required - `tool_type` string, nullable, required - `tool_call_id` string, nullable, required - `tool_description` string, nullable, required - `tool_definitions` string, nullable, required - `finish_reasons` string[], required - `error_type` string, nullable, required - `request_temperature` number, nullable, required - `request_max_tokens` integer, nullable, required - `request_top_p` number, nullable, required - `request_frequency_penalty` number, nullable, required - `request_presence_penalty` number, nullable, required - `request_seed` integer, nullable, required - `request_stop_sequences` string[], required - `request_choice_count` integer, nullable, required - `output_type` string, nullable, required - `input_messages` NormalizedMessage[], required - `role` string, required - `content` string, required - `finish_reason` string, required - `output_messages` NormalizedMessage[], required - `role` string, required - `content` string, required - `finish_reason` string, required - `system_instructions` string[], required - `tool_call_arguments` string, nullable, required - `tool_call_result` string, nullable, required - `compaction_summary` string, nullable, required - `compaction_items_before` integer, nullable, required - `compaction_items_after` integer, nullable, required - `content_refs` string[], required - `artifact_refs` string[], required - `object_refs` string[], required - `custom_attrs_string` object, required - `custom_attrs_int` object, required - `custom_attrs_float` object, required - `custom_attrs_bool` object, required - `server_address` string, nullable, required - `server_port` integer, nullable, required - `wb_user_id` string, nullable, required - `wb_run_id` string, nullable, required - `wb_run_step` integer, nullable, required - `wb_run_step_end` integer, nullable, required - `raw_span_dump` string, nullable, required - `groups` AgentSpanGroupRow[], required - `group_keys` object, required - `span_count` integer, required - `invocation_count` integer, required - `conversation_count` integer, required - `total_input_tokens` integer, required - `total_cache_creation_input_tokens` integer, required - `total_cache_read_input_tokens` integer, required - `total_output_tokens` integer, required - `total_reasoning_tokens` integer, required - `total_duration_ms` integer, required - `error_count` integer, required - `total_cost_usd` number, nullable, required - `total_input_cost_usd` number, nullable, required - `total_output_cost_usd` number, nullable, required - `agent_names` string[], required - `agent_versions` string[], required - `provider_names` string[], required - `request_models` string[], required - `conversation_names` string[], required - `first_seen` string, date-time, nullable, required - `last_seen` string, date-time, nullable, required - `first_message` AgentConversationMessagePreview, required: A truncated first/last message snippet for a grouped conversation row. `role` is the chat-timeline message type (e.g. "user_message", "assistant_message") so clients can style it consistently with the full chat view; `text` is the trimmed, length-capped preview content. - `role` 'user_message' | 'assistant_message', required - `text` string, required - `last_message` AgentConversationMessagePreview, required: A truncated first/last message snippet for a grouped conversation row. `role` is the chat-timeline message type (e.g. "user_message", "assistant_message") so clients can style it consistently with the full chat view; `text` is the trimmed, length-capped preview content. - `role` 'user_message' | 'assistant_message', required - `text` string, required - `metrics` object, required - `distributions` object, required - `total_count` integer, required ## Other responses - `422`: Validation Error ## Changes - **2026-10-22** `f7643c49bf2e`: 4 info - added the required property `spans/items/main_agent_id` to the response with the `200` status - added the required property `spans/items/main_agent_name` to the response with the `200` status - added the required property `spans/items/response_time_to_first_chunk` to the response with the `200` status - added the required property `spans/items/workflow_name` to the response with the `200` status - **2026-10-16** `dd89fadfe4e4`: 1 info - endpoint added [Operation history](https://skmtc.dev/wandb/apis/fastapi/changes/agents/spans/query/post.md) --- [API](https://skmtc.dev/wandb/apis/fastapi.md) - [All operations](https://skmtc.dev/wandb/apis/fastapi/llms.txt) - [OpenAPI document](https://skmtc.dev/wandb/apis/fastapi/revisions/f7643c49bf2e?raw)