CluesoView on GitHub

Video Skills

A skill is a ready-made set of instructions for your AI agent. These 90 teach Claude, ChatGPT, Cursor, and 20+ other agents to create and edit videos with Clueso. Install all the skills at once or browse our collection to pick specific skills to try.

How to use

  1. Open Settings → Connectors → Add custom connector (needs a paid Claude plan) and paste this URL:https://connect.clueso.io/mcp
  2. Open any skill below and hit Copy skill. No terminal, no npx.
  3. Paste it into your chat and ask for a video:Follow this skill to make a 30-second launch video for our new feature.

Skills run on the free Clueso MCP, no API keys needed. Questions? See the FAQ, or browse the collection below.

Works with

Browse the collection

Create videos

academy-course-trailer

Make a 30-45 second trailer that gets a customer to enroll in a course: name who it's for, flash three things they'll learn, land the outcome promise, and close on an enroll CTA. It sells the transforQuick edits

accessibility-pass

Audit and fix one video so it works for viewers who can't hear it, can't read small low-contrast text, or need a moment longer to follow a step - captions, legibility, and pacing in a single pass.Quick edits

add-ai-captions

Turns an existing video with spoken audio into a captioned, on-brand version: the audio is transcribed for a word-level transcript and real timing, and that timing drives placement of styled, animatedQuick edits

add-animated-captions

Most viewers watch with the sound off. This pass delivers a video with captions burned into the file itself - every spoken word readable on screen, timed to the voice, on every platform with zero playQuick edits

add-callouts-and-arrows

Turn "click the button in the corner" into an arrow that lands on that exact button as the words are spoken. Callouts and arrows do the pointing so the narration doesn't have to give directions.Quick edits

add-cta-end-card

Give a video a proper ending: one branded card, one next action, a few calm seconds to read it. No montage, no three-links-and-a-QR-code - the card exists so the viewer knows exactly what to do next.Quick edits

add-key-point-overlays

Viewers remember what they read, not just what they hear. This pass distills each section of a video into one short line and puts it on screen at the moment the narration makes that point.Quick edits

add-logo-watermark

Place the user's logo as a quiet, persistent corner mark across the whole video - visible enough to claim the content, subtle enough that viewers stop noticing it after two seconds.Quick edits

add-lower-thirds

Give a video broadcast-style wayfinding: a branded label in the lower third of the frame that slides in as each new segment begins, telling viewers where they are without interrupting anything.Quick edits

add-opening-hook

Give a video the first five seconds it deserves: a payoff-first hook that tells viewers what they'll get, replacing the cold "okay, so here's the dashboard" open. Everything after the hook stays as itQuick edits

add-zoom-emphasis

Give a flat screen recording camera direction: push in on the control being used at each key moment, hold while the action happens, and release back to full frame - so viewers always know where to looMotion graphics

animated-explainer-video

Produce the classic 60-90 second explainer: a narrated story that takes a viewer from "here's a pain you recognize" to "here's the thing that fixes it and how it works", told through generated animateDocs & articles

annotate-article-screenshots

Turn an article's raw full-window screenshots into images that do their job: each one cropped and zoomed to the region that matters, with an arrow or marker on exactly the control the step tells the rQuick edits

apply-brand-colors

Take a video whose overlays and text drifted off-brand and bring every designed element - text styles, callouts, shapes, backgrounds, title cards - onto the brand palette in a single consistent pass, Create videos

article-to-video

Turn written long-form content into a narrated explainer video: distill the text into a scene-by-scene script, compose each scene with native Clueso elements, narrate it, and export. The "intelligenceEdit videos

batch-multi-language-export

Produces N independent exports of one source video, each dubbed into a different language, from a single Clueso project. This is the production and export mechanics skill: it fans work out safely acroQuick edits

blur-sensitive-info

Make a recording safe to publish: find every piece of sensitive information on screen and keep it blurred for every frame it's visible. One leaked email in one frame defeats the whole pass, so this skMotion graphics

branded-intro-outro

Create the two bookends every team video should share: a short intro built around a logo moment with a slot for the video's title, and an outro that closes with the brand and one clear call to action.Motion graphics

brief-to-launch-video

Produce a 30-60 second launch/announcement video from a written brief. Everything - project creation, scene composition, motion, voiceover, and export - happens through Clueso. No external APIs, no loCreate videos

chaptered-training-video

Turn one long recording (onboarding walkthrough, training session, product tour) into a structured, module-by-module learning sequence: a clean split at topic boundaries, with each chapter opening on Quick edits

chapterize-video

Turn one long unbroken video into a clearly sectioned piece: a split at every topic boundary and a short, consistent title card announcing each chapter. The content itself is not cut or rewritten - itEdit videos

compliance-training-localization

Produces one export per required language from a single compliance training video: a translated script, a matching voiceover, re-synced visuals, and a rendered export per language. Mechanically this iCreate videos

course-module-video

Turn a lesson topic, outline, or SME notes into an LMS-ready course module built on instructional-design structure: objectives stated up front, a chaptered demo where every chapter maps to exactly oneCreate videos

crm-training-video

Turn a raw screen recording of someone using a CRM or internal sales tool into a clean, callout-annotated training video the rest of the team can learn from. The hard part isn't the recording: it's th

FAQ

What is an Agent Skill?

A skill is a folder with a SKILL.md file: step-by-step instructions an AI agent (Claude Code, Cursor, and 20+ others) loads on demand to perform a specialized task. These skills teach agents how to create and edit videos with Clueso.

Do I need any API keys?

No. Every skill here runs on the Clueso MCP and nothing else: no external APIs, no local binaries, no configuration. Connect the Clueso MCP in your agent and run one npx command: npx skills add clueso-ai/skills

How do I install a single skill?

Add --skill to the install command, e.g. npx skills add clueso-ai/skills --skill brief-to-launch-video. Every skill page also has a Copy skill button you can paste straight into a chat, and Claude Code users can alternatively use /plugin marketplace add clueso-ai/skills.

What is the Clueso MCP?

Clueso is an AI video platform. Its MCP server exposes video creation to agents (projects, clips, motion elements, AI voiceover, and export) so an agent can produce a finished video end to end.

Which agents are supported?

Any agent that supports Agent Skills: Claude Code, claude.ai, Cursor, Codex, Windsurf, Gemini CLI, Cline, Roo Code, Amp, opencode, Zed, GitHub Copilot, and more. The npx installer detects your agent automatically.

Can I contribute a skill?

Yes. The repository is open source under Apache-2.0. Copy the template, follow the authoring spec (declare every dependency; Clueso-MCP-only skills are strongly preferred), and open a pull request.