Full reference

Data assets and snapshots

Curate evidence from uploads or traces, record its intended use, and snapshot the exact data used by an evaluation or training run.

Updated 12 October 2026

Reviewed against Console v0.2.2. Available actions depend on project access and runtime configuration; check your installed CLI version for command differences.

Build a dataset in the current Console

Select the project, then open Lab → Datasets → New dataset.
Enter Name, choose Tag and Dates, and set Most traces. The current limit is 250, newest first.
Check the match count and inspect source content. Matching counts can include traces without captured payloads.
Create the dataset, then inspect the cases and Cleaned / In Lab counts. The builder cleans collected cases and only stages those no longer marked raw.
Keep the resulting dataset/snapshot identity with the evaluation so every candidate uses the same evidence.

Creating a dataset writes stored evidence; it does not supply human gold labels. Deterministic cleaning redacts supported patterns but does not guarantee complete anonymisation. See the evaluation walkthrough.

The current Traces screen does not expose the older Engineering → Traces → Import traces flow. For file-based evaluation assets, use the supported CLI upload below. Uploading a data asset is different from importing a provider export into the trace list.

Choose evidence with provenance

Keep the source trace or file, task context and permitted use with each asset. Separate evaluation cases from training examples when you need an independent holdout. Registering metadata does not upload the underlying data.

Upload or curate an asset

bash

stimulir data list --target eval
stimulir data upload cases.jsonl --name support-cases --target eval

Validate the file and remove sensitive content before upload. For trace curation, from-trace requires a trace ID and a supported source value; inspect its help and use evidence from the intended project.

Advance the asset lifecycle

Staging records readiness for the next operation. It does not prove that cleaning, review or evaluation has happened. Inspect the resulting state and the required fields for the target workflow. Retarget or unstage an asset deliberately when its intended use changes.

Freeze evaluation input

bash

stimulir data snapshot <asset-id>

Retain the returned snapshot ID with the run. Compare candidates on the same immutable evidence instead of a dataset that changes between runs.

Use the SDK

client.data_assets exposes creation, upload, trace curation, metadata updates, stage transitions and snapshots. The method catalogue below lists arguments. Follow Evaluation after confirming the snapshot is suitable for the task.

Command catalogue

Expand a command for its arguments. These commands follow the released CLI source; check stimulir --version and the command’s --help for your installation. Required flags are marked; some operations also validate combinations of inputs.

stimulir data bulk-stage

Stage multiple assets together for cleaning, snapshotting, Lab, SFT, or eval.

--ids required: Comma-separated data asset IDs.
--stage required: See the installed command’s --help for accepted values.
--status: See the installed command’s --help for accepted values.
--target: eval | sft | preference.
--json: Emit raw JSON output.

stimulir data create

--name, -n required: Asset display name.
--kind: See the installed command’s --help for accepted values.
--stage: See the installed command’s --help for accepted values.
--target: eval | sft | preference.
--description: See the installed command’s --help for accepted values.
--content-type: See the installed command’s --help for accepted values.
--json: Emit raw JSON output.

stimulir data from-trace

Curate a data asset directly from a trace.

trace_id required: Positional argument. Trace UUID (from stimulir lab eval ... or the traces panel).
--source required: See the installed command’s --help for accepted values.
--name, -n: See the installed command’s --help for accepted values.
--target: See the installed command’s --help for accepted values.
--json: Emit raw JSON output.

stimulir data list

List data assets in the active workspace.

--cursor: Pagination cursor.
--limit: See the installed command’s --help for accepted values.
--target: Filter by eval | sft | preference.
--json: Emit raw JSON output.

stimulir data remove

Remove a data asset (irreversible).

asset_id required: Positional argument. Asset ID.
--yes, -y: Skip the confirmation prompt.
--json: Emit raw JSON output.

stimulir data snapshot

Snapshot the current state of an asset (immutable point-in-time copy).

asset_id required: Positional argument. Asset ID.
--json: Emit raw JSON output.

stimulir data stage

Transition an asset's lifecycle stage (raw → cleaning → … → lab).

asset_id required: Positional argument. Asset ID.
stage required: Positional argument. See the installed command’s --help for accepted values.
--status: See the installed command’s --help for accepted values.
--json: Emit raw JSON output.

stimulir data unstage

Move a staged asset back to raw so it can be cleaned or re-targeted.

asset_id required: Positional argument. Asset ID.
--json: Emit raw JSON output.

stimulir data update

Update data asset metadata without changing the underlying stored payload.

asset_id required: Positional argument. Asset ID.
--name, -n: Asset display name.
--description: See the installed command’s --help for accepted values.
--target: eval | sft | preference.
--status: See the installed command’s --help for accepted values.
--json: Emit raw JSON output.

stimulir data upload

Upload a file as a new data asset.

file required: Positional argument. File to upload.
--name, -n: Override the asset name.
--stage: See the installed command’s --help for accepted values.
--target: See the installed command’s --help for accepted values.
--description: See the installed command’s --help for accepted values.
--json: Emit raw JSON output.

SDK method catalogue

Expand a client to inspect public methods and parameter names. Read each operation’s task guide before supplying a payload or starting work. The SDK returns platform responses; handle errors and inspect completion state before using a result.

client.data_assetsDataAssetsClient

list(stage, kind, status, target, tag, limit): Returns the platform response; inspect its status and error detail.
create(name, kind, stage, target, tags): Create a new data asset.
upload(file, name, target): Returns the platform response; inspect its status and error detail.
from_trace(trace_id, source, name, target, payload, tags): Create a data asset from an existing trace.
patch(asset_id): Returns the platform response; inspect its status and error detail.
stage(asset_id, stage, status, lineage_event): Returns the platform response; inspect its status and error detail.
unstage(asset_id): Returns the platform response; inspect its status and error detail.
bulk_stage(asset_ids, stage, status, target, lineage_event): Returns the platform response; inspect its status and error detail.
snapshot(asset_id): Returns the platform response; inspect its status and error detail.