SkillNarrative Audience

Create Lookalike Audience

Build a look-alike audience by scoring a candidate population against a seed and materializing the matches.

About

Overview

Create a look-alike audience from a seed audience and a candidate population dataset. Classifies Rosetta Stone attributes, generates the same materialized-view scoring pipeline Lookalike Studio emits (Naive-Bayes categorical weights + Gaussian continuous similarity), gates on approval, submits via a workflow, and monitors the build to completion.

AudienceLookalikeWorkflowScoringActivation

Skill Support

Specifications

Plugin: Narrative Audience
Runtime: Prompt + code
Version: 0.1.0
Languages: NQL, YAML
Required Tools: Read, narrative_context_get, narrative_datasets_search, narrative_datasets_describe, narrative_dataset_get_column_stats, narrative_attributes_search, narrative_nql_validate, narrative_data_planes_list, narrative_workflows_create, narrative_workflow_runs_list, narrative_jobs_search, narrative_jobs_describe
Included Files: SKILL.md, scripts/lookalike_state_tag.py

Arguments

Argument	Required / Default	Description
`--seed <id\|name>`	Optional	The seed audience dataset — the users the output should look like. Numeric ID or datasetName. If omitted, the skill asks.
`--population <id\|name>`	Optional	The candidate population dataset to score and select from. Numeric ID or datasetName. If omitted, the skill asks.
`--size <N>`	Optional	Output mode "size": keep the top N highest-scoring candidates. Mutually exclusive with --min-score.
`--min-score <0..1>`	Optional	Output mode "score": keep every candidate whose similarity probability is at or above this threshold (clamped to 0.001–0.999 and converted to log-odds internally). Mutually exclusive with --size.
`--include-seed`	Default: false	Union the seed members back into the output audience with a perfect score of 1.0. Off by default — the output is look-alikes only.
`--name <audience name>`	Optional	Unique name for the output audience dataset. Slugified to form the pipeline prefix. Must not collide with an existing dataset.
`--data-plane <id>`	Optional	UUID of the data plane to target. Skips data-plane resolution.
`--no-trigger`	Optional	Create the workflow without triggering a run. Default is trigger_immediately: true — the audience builds right away.
`--no-state-tag`	Optional	Skip the _nio_lookalike_serialization wizard-state tag on the output audience. Default is to include it so Lookalike Studio can reopen the audience for editing.
`--dry-run`	Optional	Render and explain the full pipeline but do NOT submit. Implies --show-spec.
`--show-spec`	Optional	Include the full rendered workflow YAML in the approval preview. Off by default; the plain-English summary is enough for most users.
`<free-text tail>`	Optional	The user's intent (e.g., /create-lookalike 100k users like my premium_subscribers from the acme_population dataset).

Example prompts

Prompt	When to use it
`Create a lookalike audience from my premium subscribers`	Score a candidate population against a seed segment and keep the closest matches.
`Expand my seed audience to 500k similar users`	Use size mode to keep the top N highest-scoring look-alike candidates.
`Score the population against my customers`	Build the Naive-Bayes + Gaussian similarity pipeline and materialize the scored audience.
`/create-lookalike 100k users like my premium_subscribers from the acme_population dataset`	Bind seed and population in one line, build, and trigger the audience workflow immediately.