Rank Moments

Purpose

Rank discovered moments into final Sparks for review or auto-render.

Install from SKLx

This skill is published on SKLx and can be installed locally or on an Opal server:

skl add "rank moments" --registry https://sklx.one

Current API

Ranking is an internal job phase. Poll the job to inspect progress.

GET /api/jobs/:jobId

Inputs

{
  "jobId": "string",
  "candidates": []
}

Outputs

{
  "jobId": "string",
  "status": "ready_for_review",
  "clips": []
}

Steps

1. Load candidate moments.
2. Rank by hook, clarity, shareability, standalone meaning, pacing, and payoff.
3. Select strongest Sparks.
4. Polish captions.
5. Persist selected Sparks on the job.

Rules

Do not invent timestamps. Avoid duplicate ideas. Prefer clips with clear payoff. Keep only strong standalone moments. Return valid JSON from model calls.