inject-scenarioslisted

# Inject scenarios Use `rockfish.labs.scenarios.Client` to inject synthetic scenarios into a time-series dataset and optionally generate Q&A test cases from the perturbation. ## When to use this skill Use when the user has (or will create) a baseline time-series dataset and wants to: - Inject a **spike** at a single timestamp (set `measurement = magnitude`). - Inject an **outage** across a time range (hold `measurement = outage_value` between `start_timestamp` and `end_timestamp`). - Inject a **ramp** or **shift** (gradual or step change). - Generate Q&A test cases describing the injected scenario. - List every scenario derived from a source dataset. If the user doesn't yet have a baseline dataset, use the `generate-from-schema` skill first to produce one. ## Concept A *scenario* is a perturbation applied to a single `measurement` column in a time-series dataset, indexed by a `timestamp_column`. The scenarios service: 1. Reads the source dataset by `dataset_id`. 2. Applies the perturbation per the `config` payload. 3. Writes a new dataset containing the perturbed series, with lineage labels (`source_dataset_ids`) and scenario metadata (`scenario_config`) stamped onto the Arrow schema. ## How to use ```python from rockfish.labs.scenarios import Client async with rf.Connection.from_config() as conn: client = Client(conn) result = await client.inject( source_id, config={ "type": "spike", # or outage, ramp, shi