# Read one step of a session's trace in full: tool input and output, the model request and response, and its raw events. Source: GET /v1/sessions/{session\_id}/trace/turns/{turn}/steps/{step} Select code samplecURL (shell:curl) ```js const url = 'https://app.everruns.com/api/v1/sessions/example/trace/turns/example/steps/example'; const options = {method: 'GET'}; try { const response = await fetch(url, options); const data = await response.json(); console.log(data); } catch (error) { console.error(error); } ``` ```sh curl --request GET \ --url https://app.everruns.com/api/v1/sessions/example/trace/turns/example/steps/example ``` * Production APIapp.everruns.com/api/v1/sessions/{session\_id}/trace/turns/{turn}/steps/{step} ## Parameters ### Path Parameters **session\_id** required string Prefixed public identifier **turn** required string Prefixed public identifier **step** required string Prefixed public identifier ### Query Parameters **full** boolean Return payloads whole, however large. ## Responses ### 200 Success Media typeapplication/json Everything about one step, for the inspector. object **events** required Raw events of the step, in sequence order. Array\ An event in the session log. object **data** Payload, when small enough to inline. **id** required Event identifier. string **sequence** required Sequence in the session event log. integer format: int32 **size\_bytes** required Size of the serialized payload. integer format: int32 **ts** required When the event happened. string format: date-time **type** required Event type, such as `tool.call_completed`. string **input** One of: * object Tool arguments. object **preview** The first part of the serialized value, when it was cut. string | null **size\_bytes** required Size of the serialized value. integer **truncated** required Whether the value was cut to `preview`. boolean **value** The value, when it fits or `full` was asked for. * null null **output** One of: * object Tool result, or error. object **preview** The first part of the serialized value, when it was cut. string | null **size\_bytes** required Size of the serialized value. integer **truncated** required Whether the value was cut to `preview`. boolean **value** The value, when it fits or `full` was asked for. * null null **request** One of: * object The request of a model call: what changed since the previous call. object **message\_count** required Messages sent in the call. integer format: int32 **new\_from** required Messages before this index were already in the previous call. integer format: int32 **new\_messages** required The new messages, at most 50. Array\ One message of a model request. object **content** The whole message, when asked for and within the limit. **index** required Position in the request. integer format: int32 **preview** required Start of the message text. string **role** required Role as handed to the provider driver: `system`, `user`, `assistant` or `tool`. Stored messages say `agent` and `tool_result`; each driver then maps these four onto its own wire format. string **size\_bytes** required Size of the serialized message. integer **system\_preview** Start of the system prompt. string | null **tool\_count** required Tools offered in the call. integer format: int32 **usage** One of: * object Token usage of the call, split by prompt cache. object **cache\_creation\_tokens** required Prompt tokens written to the provider’s cache. integer format: int64 **cache\_read\_tokens** required Prompt tokens read from the provider’s cache. integer format: int64 **cache\_reset** One of: * object Set when this call lost most of the prompt prefix the previous model call had cached. object **idle\_seconds** required Seconds between the end of the previous call and the start of this one. integer format: int64 **lost\_tokens** required How many of those this call did not read back. integer format: int64 **previous\_cached\_tokens** required Prompt tokens the previous call left in the cache (read plus written). integer format: int64 **reason** required Likely cause, the first difference found: `model_changed`, `tools_changed`, `system_changed`, `history_changed`, `expired` (idle longer than the cache retention), or `unknown`. string * null null **cached\_through** Estimated index of the last message inside the cached prefix: the cache-read share of the prompt laid over the messages by size. `-1` when only the tools fit; absent when nothing was read from cache. integer | null format: int32 **caching\_would\_help** A long prompt sent with caching off that neither read nor wrote the cache: turning on prompt caching would likely make calls like it cheaper. boolean **input\_tokens** required Prompt tokens neither read from nor written to the cache. integer format: int64 **output\_tokens** required Completion tokens. integer format: int64 **prompt\_cache** How prompt caching was requested, such as `auto` or `anthropic_cache_control`; absent when caching was off. string | null * null Set when this call lost most of the prompt prefix the previous model call had cached. object **idle\_seconds** required Seconds between the end of the previous call and the start of this one. integer format: int64 **lost\_tokens** required How many of those this call did not read back. integer format: int64 **previous\_cached\_tokens** required Prompt tokens the previous call left in the cache (read plus written). integer format: int64 **reason** required Likely cause, the first difference found: `model_changed`, `tools_changed`, `system_changed`, `history_changed`, `expired` (idle longer than the cache retention), or `unknown`. string * object null * null null * null Token usage of the call, split by prompt cache. object **cache\_creation\_tokens** required Prompt tokens written to the provider’s cache. integer format: int64 **cache\_read\_tokens** required Prompt tokens read from the provider’s cache. integer format: int64 **cache\_reset** One of: * object Set when this call lost most of the prompt prefix the previous model call had cached. object **idle\_seconds** required Seconds between the end of the previous call and the start of this one. integer format: int64 **lost\_tokens** required How many of those this call did not read back. integer format: int64 **previous\_cached\_tokens** required Prompt tokens the previous call left in the cache (read plus written). integer format: int64 **reason** required Likely cause, the first difference found: `model_changed`, `tools_changed`, `system_changed`, `history_changed`, `expired` (idle longer than the cache retention), or `unknown`. string * null null **cached\_through** Estimated index of the last message inside the cached prefix: the cache-read share of the prompt laid over the messages by size. `-1` when only the tools fit; absent when nothing was read from cache. integer | null format: int32 **caching\_would\_help** A long prompt sent with caching off that neither read nor wrote the cache: turning on prompt caching would likely make calls like it cheaper. boolean **input\_tokens** required Prompt tokens neither read from nor written to the cache. integer format: int64 **output\_tokens** required Completion tokens. integer format: int64 **prompt\_cache** How prompt caching was requested, such as `auto` or `anthropic_cache_control`; absent when caching was off. string | null * object Set when this call lost most of the prompt prefix the previous model call had cached. object **idle\_seconds** required Seconds between the end of the previous call and the start of this one. integer format: int64 **lost\_tokens** required How many of those this call did not read back. integer format: int64 **previous\_cached\_tokens** required Prompt tokens the previous call left in the cache (read plus written). integer format: int64 **reason** required Likely cause, the first difference found: `model_changed`, `tools_changed`, `system_changed`, `history_changed`, `expired` (idle longer than the cache retention), or `unknown`. string * null null * object null * null null **step** required The step itself. object **child\_session\_id** Session a `spawn_agent` call started. string | null **duration\_ms** Wall time of the step; absent while running. integer | null format: int64 **end\_sequence** Last event sequence of the step. integer | null format: int32 **input\_tokens** Prompt tokens of a model call. integer | null format: int32 **kind** required `model`, `answer`, `tool`, `approval`, `agent` or `send`. string **model** Model that served the call, for model steps. string | null **name** Tool name, or model for model calls. string | null **narration** What the model said alongside this call (model calls and answers). string | null **offset\_ms** required Start relative to the turn’s start. integer format: int64 **output\_tokens** Completion tokens of a model call. integer | null format: int32 **requested\_tool\_call\_ids** Tool calls this model call asked for. Array\ **result** Short result or error. string | null **start\_sequence** required First event sequence of the step. integer format: int32 **started\_at** required When the step started. string format: date-time **status** required `running`, `success`, `error` or `cancelled`. string **step** required Per-turn step number, 1-based. integer format: int32 **target** What the call acted on, from the tool’s narration. string | null **tool\_call\_id** Tool call identifier, for tool steps. string | null **turn** required Turn the step belongs to. integer format: int32 ##### Example ```json { "events": [ { "id": "event_01a124cd8a1b7f02a5c4e6d1b2f3a4c5", "sequence": 14, "size_bytes": 512, "type": "tool.started" } ], "input": { "size_bytes": 221, "truncated": false }, "output": { "size_bytes": 221, "truncated": false }, "request": { "message_count": 14, "new_from": 12, "new_messages": [ { "index": 12, "role": "tool", "size_bytes": 244 } ], "tool_count": 9, "usage": { "cache_creation_tokens": 0, "cache_read_tokens": 12529, "cache_reset": { "idle_seconds": 42, "lost_tokens": 18420, "previous_cached_tokens": 18420, "reason": "tools_changed" }, "cached_through": 11, "input_tokens": 10295, "output_tokens": 197, "prompt_cache": "auto" } }, "step": { "duration_ms": 269, "end_sequence": 21, "kind": "tool", "model": "gpt-6.1-sol", "name": "bash", "offset_ms": 3010, "start_sequence": 14, "status": "success", "step": 2, "tool_call_id": "call_PzDcvnFaUll8os3tPhWj", "turn": 1240 } } ```