Pexo Launch Video Agent Skill for Openclaw

A specialized AI agent skill that communicates with the Pexo platform to draft scripts, generate scenes, and render publish-ready launch videos for startups and products.

pexo
v0.1.0
Jun 11, 2026
0
428
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install launch-video

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install launch-video using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Pexo Launch Video Agent Skill?

The Pexo Launch Video skill is a powerful addition to your AI workflow, allowing autonomous agents to generate professional launch videos directly from a founder's pitch, landing page URL, or screenshot. Rather than attempting to generate heavy video assets locally, this skill implements a precise relay-and-poll mechanism to coordinate with Pexo's specialized cloud rendering engines.

By incorporating this capability into your suite of Openclaw Skills, your AI coding agent can write scripts, select cinematic models, sync background music, and assemble a final high-resolution video ready for Product Hunt, social media, or landing pages. The core philosophy of this skill is delegation: the local AI agent acts as a reliable pipeline coordinator, relaying verbatim user requests and assets to Pexo and handling state-based feedback loops until the complete video is delivered.

Pexo Launch Video Agent Skill Use Cases

  • Startup Launch Campaigns: Rapidly create high-conversion launch videos for Product Hunt, Twitter/X, and landing pages.
  • Product Teaser Generation: Turn landing page mockups or screenshots into dynamic, fully edited product teasers.
  • Continuous Revision Iteration: Tweak existing video drafts by supplying incremental feedback without losing original context, all handled via Openclaw Skills integration.
  • Automated Workflow Integration: Let autonomous software development agents generate marketing material as part of a continuous deployment pipeline.

How Pexo Launch Video Agent Skill Works

  1. Project Creation: The agent initiates a project using pexo-project-create.sh and obtains a unique project_id.
  2. Asset Upload: If the user provides mockups or audio, the agent uploads them via pexo-upload.sh and notes the returned tags.
  3. Request Dispatch: The agent sends the exact prompt to Pexo using pexo-chat.sh, maintaining strict fidelity to the user's instructions.
  4. Polling & Status Monitoring: The agent enters a polling loop using pexo-project-get.sh to track the rendering progress and interpret the nextAction state machine (WAIT, RESPOND, DELIVER, FAILED, RECONNECT).
  5. Interactive Selection: During the RESPOND state, the agent presents preview variations to the user, grabs their selection, and submits it back to the pipeline.
  6. Final Delivery: Upon reaching the DELIVER state, the agent requests the high-resolution asset URL and presents it unformatted to the user.

Pexo Launch Video Agent Skill Setup

Configure your Pexo API credentials locally to enable the skill within your agent workspace.

Prerequisites

Ensure your system has curl, jq, and file installed and accessible in your shell environment.

Configuration

Create or update the configuration file at ~/.pexo/config with your credentials:

PEXO_BASE_URL="https://pexo.ai"
PEXO_API_KEY="sk-<your-api-key>"

To verify your integration setup and run diagnostics, execute:

./scripts/pexo-doctor.sh

If you do not have an active account, consult references/SETUP-CHECKLIST.md for signup details and promotional credit codes.

Pexo Launch Video Agent Skill Data Schema & Taxonomy

The skill maintains state dynamically through file identifiers and project sessions on the host platform. Here is the layout of how operations mapped via Openclaw Skills align with Pexo's schema:

Command / Action Inputs Required Outputs / Saved Metadata
Project Initialization Short text brief project_id
Asset Management Local file path, Project ID asset_id (referenced via tags like <original-image>)
Chat & Dispatch Project ID, Verbatim prompt, asset tags Async execution confirmation
Polling Loop Project ID JSON payload containing nextAction and recentMessages
Asset Retrieval Project ID, Asset ID JSON object containing the absolute download url

All session contexts are saved server-side by Pexo, allowing agents to issue revision queries against the exact same project_id without resetting the generation history.

Pexo Launch Video Agent Skill Advanced Features

  • State-Driven Workflow Polling: Sophisticated state management supporting interactive feedback loops (such as choosing between previews A, B, or C).
  • Resilient Auto-Reconnection: Detects aborted network sessions and uses the RECONNECT flag to issue a "continue" command to prevent generation loss.
  • Asset Tagging Injector: Automatically maps uploaded assets like logos or screenshots using proprietary XML-style tags to maximize generation context within Pexo.
  • Credit Recovery Prompts: Intelligently identifies credit exhaustion errors and surfaces secure checkout links to the operator.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*