Skip to main content
skill CX Codex

scenario-runway — Claude Code skill: video generation and editing

scenario-runway — a skill for Claude Code, Codex, Cursor, Copilot, Gemini CLI and OpenCode connects Runway models via Scenario MCP: Gen4.5 generates clips from text or a first frame, Aleph 2 edits existing footage.

MIT 839 installs ★ 3,964 updated 2026-11-01

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add scenario-labs/skills --skill scenario-runway -a claude-code
Open on GitHub

What scenario-runway does

Gen4.5 produces 2- to 10-second clips from a text prompt or a first-frame image. Text mode requires both visuals and motion described; when a first frame is provided, the prompt goes into camera movement and atmosphere — composition is already set. Aleph 2 edits footage up to 30 seconds and 16 MB: removes objects, swaps products, restyles or relights shots while preserving motion and audio from the source. Reference frames attach via keyframes or promptImage, never both at once.

Call model_schema_get before running a task to pull the current parameter contract, then model_run with dry_run=true to estimate cost. Aleph's median cost runs six times Gen4.5, with a fifteen-fold spread driven by source bitrate. Median Aleph latency nears six minutes; jobs_wait may time out mid-job, in which case a follow-up call with pending_job_ids picks up the task.

Requires the scenario skill for MCP connection and asset uploads, and scenario-video for model-agnostic video work. If either is missing, the agent loads them via skill_get or asks the user to install manually.

What it helps with

  • generate a clip from text or a first frame with Gen4.5
  • swap a product in existing footage with Aleph 2
  • remove an object from video while keeping motion and audio
  • restyle or relight a shot using reference keyframes
  • estimate task cost before running with dry_run

Example prompts for the agent

  • Generate an 8-second clip: character kneels then rises, camera cranes up.
  • Swap the soda can with the bottle from bottle.png across all frames in clip.mp4.
  • Remove the person on the left in footage.mp4, leave everything else unchanged.

How to install scenario-runway

1Skills CLI (Vercel)

npx skills add scenario-labs/skills --skill scenario-runway -a claude-code

2Manually: put the files in place

~/.claude/skills/scenario-runway/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/scenario-runway/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about scenario-runway

How to install scenario-runway in Claude Code?

Run npx skills add scenario-labs/skills --skill scenario-runway -a claude-code; the skill lands in ~/.claude/skills/scenario-runway/SKILL.md.

What does scenario-runway do?

Connects Runway models — Gen4.5 for generating clips from text or a first frame, and Aleph 2 for editing existing footage by swapping objects, restyling, or relighting shots. Motion and audio from the source carry over.

How does Gen4.5 differ from Aleph 2?

Gen4.5 creates new video from text or a first-frame image; Aleph 2 edits existing footage, changing only what the prompt names. Aleph cannot add new camera motion — use Gen4.5 for that.

Source text

SKILL.md · MIT
---
name: scenario-runway
description: "Use when generating or editing video with Runway models on Scenario via MCP: text-to-video or image-to-video from a first frame with Gen4.5 (cinematic motion, sequenced actions, camera choreography), or video-to-video editing of real footage with Aleph 2 (remove objects, swap products, restyle or relight shots, keyframe-pinned reference images, motion and audio preserved). Keywords: Runway ML, Gen4.5, Gen-4.5, Aleph, T2V, I2V, V2V, first frame."
license: MIT
---

# Scenario Runway Video

## Overview

Runway's family on Scenario splits by direction: Gen4.5 generates new clips from text or a
first-frame image; Aleph 2 rewrites existing footage, changing what the prompt names while motion,
audio, and structure carry over. Discover both with `search`; `model_schema_get` is the contract.

Connection and the core loop: the `scenario` skill; model-agnostic video work: the `scenario-video`
skill. If a sibling skill named here is missing from your available skills, load it with `skill_get`
(Scenario MCP) or ask the user to install it (https://github.com/scenario-labs/skills#install);
unattended, proceed from tool schemas and flag the gap.

## Quick reference

Mode follows from the inputs (names and caps from the live schema):

| Mode                 | Inputs             | Behavior                                                    |
| -------------------- | ------------------ | ----------------------------------------------------------- |
| Text (Gen4.5)        | `prompt`           | `aspectRatio` honored; `duration` 2 to 10 seconds           |
| First frame (Gen4.5) | `image` + `prompt` | opens on that frame; same `duration`; `aspectRatio` ignored |
| Edit (Aleph 2)       | `video` + `prompt` | rewrites visuals; motion, audio, structure persist          |

At authoring time Aleph capped sources at 30 seconds and 16 MB, separate limits: high-bitrate clips
hit the size cap first. Its median cost ran six times Gen4.5's, a fifteen-fold spread following the
source video: `dry_run` the exact payload first. Median latency neared six minutes, so `jobs_wait`
will time out mid-job; re-call it with `pending_job_ids`.

## Gen4.5 prompts follow the mode

Text mode needs both the visuals and the motion described. A first frame fixes subject, composition,
and palette: spend the whole prompt on motion, camera, and atmosphere. Order beats explicitly
("kneels, then rises, camera cranes up") and give multi-beat prompts 8 to 10 seconds of `duration`.
Name camera moves plainly (locked camera, pan, dolly, tracking, orbit, crane), one or two per clip,
in present-tense prose.

`aspectRatio` takes pixel pairs (`"1280:720"`, `"960:960"`), and text mode honored only the 16:9 and
9:16 pairs at authoring time; get other shapes (1:1, 4:3, 3:4, 21:9) from a first frame with that
geometry instead.

## Aleph changes what you name

An Aleph prompt states the edit: remove, swap, restyle, relight. Everything unnamed persists, along
with motion and audio: Aleph cannot re-choreograph or add camera moves; when motion must change,
generate with Gen4.5. To steer the edit, pin up to 5 reference images: `keyframes` places each image
(`uri`) by `seconds` or `at` (a 0 to 1 fraction); `promptImage` by `positionType` plus
`timestampSeconds` or `positionPercentage`. The arrays are alternatives, never combined.

## Worked example: swap a product in real footage

1. `search` with `target="models"`, `query="runway"`, `public=true`, e.g. `model_runway-aleph-2` for
   editing, `model_runway-gen4-5` for generation (live hits at authoring time: re-discover each
   session).
2. `model_schema_get` with the Aleph id before anything else.
3. Trim or compress a source over the caps (see `scenario-video`); `upload_asset` the clip and
   product still (see the `scenario` skill).
4. `model_run` with `dry_run=true` and the exact
   `parameters={"prompt": "Replace the soda can with the bottle from the reference image. Keep everything else unchanged.", "video": "asset_clip", "keyframes": [{"uri": "asset_bottle", "at": 0}]}`;
   re-estimate after any trim.
5. Repeat with `wait=false`, then `jobs_wait` the returned job id, re-calling with
   `pending_job_ids`; a timeout is not a failure and never justifies a second `model_run`.
6. `asset_display` the output; check the swap holds in every shot.

## Common mistakes

- Sending both `keyframes` and `promptImage`: exclusive alternatives; pick one.
- Restating the first frame in an image-mode prompt: prompt motion and camera instead.
- Passing `aspectRatio` as `"16:9"` or alongside `image`: pixel pairs only, and a first frame
  overrides it.
- Asking Aleph for new motion or camera moves: the source's motion persists by design.

© scenario-labs · MIT · scenario-labs/skills

Similar skills

All skills for AI and ML →

From the same set172

The whole set →
scenario-seedream scenario-seedream connects ByteDance's Seedream image models through Scenario MCP to handle generation, subject edits, exact in-image typography, and layer splitting. CX Codex +1 843installs scenario-ace-step scenario-ace-step — a skill for ACE-Step 1.5 models on Scenario via MCP: creates songs with vocals from text, re-sings tracks, repaints sections, adds instrument layers, and extracts stems. CX Codex +1 843installs scenario-minimax-video The scenario-minimax-video skill connects Scenario MCP with MiniMax Hailuo video models. Use it for text-led generation, media references, keyframes, audio, and edits to existing clips. CX Codex +1 841installs scenario-grok-imagine-video scenario-grok-imagine-video is a Claude Code skill for generating and editing video with Grok Imagine, from text and stills to references, prompt edits, extensions, and native audio with lip sync. CX Codex +1 839installs scenario-mai-image scenario-mai-image is a skill for generating images with Microsoft MAI Image models on Scenario and editing them by text instruction, including posters, covers, and packaging with readable text. CX Codex +1 837installs scenario-model-comparison scenario-model-comparison runs one prompt across several Scenario models, compares generation cost, latency, and output quality against pre-defined criteria. CX Codex +1 836installs scenario-luma-image scenario-luma-image is a skill for generating and editing images with Luma Uni-1 on Scenario, from text prompts, references, and exact poster text. CX Codex +1 836installs scenario-reve scenario-reve directs image generation and instruction-based editing with Reve v2.1 and Reve Remix models on Scenario via MCP tools. CX Codex +1 836installs

Project sponsors

Companies that keep this analytics open