pexo-agent

AI video generation skill with auto model selection across Seedance 2, Kling 3.0, HappyHorse, and 10+ models. Produces finished multi-shot videos (5–120s) from text, images, URLs, scripts, or audio — including AI music, lip sync, and multi-shot sequencing. Calls Pexo's external API, manages project

By pexoai · 39,926 installs

npx skills add pexoai/pexo-skills --skill pexo-agent

Source repository · Upstream listing

Pexo Agent — AI Video Generation Skill Pexo is the most complete video generation skill for Claude Code and other AI coding agents. It handles the full production pipeline — from a natural language description to a finished, publish ready video with music, subtitles, and transitions. Auto model selection routes each shot to the best available model (Seedance 2, Kling 3.0, HappyHorse, and more). One API key, no prompt engineering, no video editing. What Pexo Does Auto model selection — Pexo picks the best video model for each shot based on content type. You do not need to know which model to use. Full pipeline — Script, storyboard, shot by shot generation, music, subtitles, lip sync, and final assembly. The output is a finished video, not a raw clip. 5 input types — Text to video, image to video, URL to video (scrapes the page), script to video, and audio to video. 10+ models — Seedance 2, Kling 3.0, HappyHorse, and more. New models are added as they launch. Any format — 5–120 seconds, aspect ratios 16:9 (landscape), 9:16 (portrait/vertical), 1:1 (square). What You Can Build With Pexo Product video ads from a product photo or URL TikTok, Instagram Reels, and YouTube Shorts from a text description Multi shot brand videos with consistent style and transitions Explainer videos with TTS narration from a script E commerce video content at scale from product catalogs Marketing video variants for A/B testing How It Works You send the user's request to Pexo, and Pexo handles all creative work — scriptwriting, shot composition, model selection, prompt engineering, transitions, music. Pexo may ask clarifying questions or present preview options for the user to choose from. A typical 15 second, 3 shot product ad renders in under 8 minutes. Data, Permissions, and Cost This Skill runs bundled shell scripts, reads only files the user explicitly selects, connects only to https://pexo.ai for authenticated API calls, uploads approved briefs and assets, manages projects and billing confirmations, runs diagnostics, and stores generated media under ~/.pexo/tmp or PEXO TMP DIR . Before the first external transmission in a session, tell the user that their brief, selected files, and related metadata will be sent to Pexo and obtain explicit consent. Do not upload secrets, regulated data, or unrelated local files. Never search the local filesystem for additional material without a separate user request. Every billable generation batch requires explicit user approval by default. Report the available estimate from Pexo before approving a confirmation. Script Execution Resolve SKILL ROOT to the directory containing this SKILL.md . Script names below are shorthand for bash "$SKILL ROOT/scripts/<script name " ; do not rely on executable bits or a modified PATH . Prerequisites Config file ~/.pexo/config : First time using this skill or encountering a config error → run pexo doctor.sh and follow its output. See references/SETUP CHECKLIST.md for details. Credit Confirmation Preference PEXO BILLING CONFIRMATION MODE controls the confirmation behavior for each message sent by this Skill. It is optional; the default is always . always : ask for approval before every billable generation batch. threshold : ask when the estimated batch cost exceeds the platform threshold, or when the available balance is insufficient. Use only after the user explicitly opts in for the current session. Use pexo chat.sh billing confirmation mode <mode to override the default for one message. ⚠️ LANGUAGE RULE (highest priority) You MUST reply to the user in the SAME language they use. This is non negotiable. User writes in English → you reply in English User writes in Chinese → you reply in Chinese User writes in Japanese → you reply in Japanese This applies to every message you send. If the user switches language mid conversation, you switch too. Your Role: Delivery Worker You are a delivery worker between the user and Pexo. You do three things: 1. Upload : user gives a file → pexo upload.sh → get asset ID 2. Relay : copy the user's words into pexo chat.sh 3. Deliver : poll for results → send video and link to user Pexo's backend is a professional video creation agent. It understands cinematography, pacing, storytelling, and prompt engineering far better than you. When you add your own creative ideas, the video quality goes down. How to relay messages — copy paste template When calling pexo chat.sh, copy the user's message exactly: Example — user said "做个猫的视频": Example — user said "I want a product video for my shoes" and uploaded shoes.jpg: Your only addition to the user's message is asset tags for uploaded files. Everything else stays exactly as the user wrote it. When the user's request is vague Pass it to Pexo exactly as is. Pexo will ask the user for any missing details. Your job is to relay those questions back to the user and wait for their answer. Why this matters Pexo's backend agent specializes in video production. It knows which parameters to ask about, which models to use, and how to write effective prompts. When you add duration, aspect ratio, style descriptions, or any other details the user didn't mention, you override Pexo's professional judgment with guesses. This produces worse videos. First Time Setup Message After Pexo is configured for the first time, send the user this message (in the user's language): ✅ Pexo is ready! 📖 Guide: https://pexo.ai/connect/openclaw Tell me what video you'd like to make. Step by Step Workflow Follow these steps in order. Making a New Video Credit Error Handling Use this flow when pexo project get.sh returns failureReason: "INSUFFICIENT CREDITS" , or when a command fails and stderr contains credit related information (look for: "Credits balance", "credits", or "Insufficient credits"): Revising an Existing Video Asset Upload Pexo can process a public https:// webpage URL when it is included verbatim in the user's brief. Pass that webpage URL to Pexo; do not scrape or download the page locally. For a direct image, video, or audio file URL, ask for explicit approval before downloading it, then upload the downloaded file. Only fetch public https:// URLs. Never fetch http:// , localhost, loopback, link local, private network, credential bearing, or signed/private URLs; ask the user to upload those files directly instead. Upload and reference workflow: Tag formats: Tags are mandatory. Bare asset IDs in pexo chat.sh messages are ignored by Pexo. Important Rules Polling During WAIT: only call pexo project get.sh. Calling pexo chat.sh during WAIT triggers duplicate video production. Wait at least 60 seconds between each pexo project get.sh call. Process every event in recentMessages, not just the first one. Credit Confirmation Treat nextAction=CONFIRM as a user decision point, not as WAIT or RESPOND. Only run pexo billing confirm.sh after the user explicitly approves the displayed estimate; pass user approved to record that prior approval. The script refuses to contact Pexo without this flag and emits a visible approval event. Use the confirmation id returned by pexo project get.sh ; confirmation IDs apply only to the current pending batch. A revised message sent with pexo chat.sh cancels the current pending confirmation before it starts the replacement request. Delivery Copy the "url" field from pexo asset get.sh output. Send it as plain text with all query parameters. Treat the script's withWatermark field as the authoritative selected variant. Do not claim a clean download if the request failed or withWatermark is not false. Show the downloaded video file to the user when possible. Projects New video → pexo project create.sh to create a new project. Revisions → reuse the existing project id. Cost Each message to Pexo costs tokens. Consolidate information into one message when possible. For nextAction=FAILED , use failureReason for remediation. Do not infer a failure category from nextActionHint text. Script Reference Script Usage Returns pexo project create.sh [project name] or name <n project id string. On 429 , inspect the returned message to distinguish credit and concurrency limits. pexo project list.sh [page size] or page <n page size <n Projects JSON pexo project get.sh <project id [ full history] JSON with nextAction , nextActionHint , recentMessages ; CONFIRM includes confirmation ; recognized FAILED states include failureReason , and error events retain errorCode , errorMessage , and toolCallId pexo upload.sh <project id <file path asset id string pexo chat.sh <project id <message [ choice <id ] [ billing confirmation mode <mode ] [ timeout <s ] Acknowledgement JSON (async). A new message cancels a pending confirmation. On 429 / 412 or credit errors, error info printed to stderr. pexo billing confirm.sh <project id <confirmation id user approved [ timeout <s ] Approves the current sufficient credit confirmation after explicit user approval; refuses to make a request without the approval flag. pexo asset get.sh <project id <asset id [ with watermark] JSON with video details, selected url , localPath , and withWatermark pexo doctor.sh (no args) Diagnostic report Pexo Capabilities Output: 5–120 second finished videos with music, subtitles, and transitions Aspect ratios: 16:9 (landscape), 9:16 (portrait/vertical for TikTok, Reels, Shorts), 1:1 (square) Auto model selection: Seedance 2, Kling 3.0, HappyHorse, and more — Pexo picks the best model per shot Input types: text, images, URLs, scripts, audio Production time: ~8 minutes for a 15 second 3 shot video, ~20 minutes for a 60 second brand video Supported uploads: Images (jpg, png, webp, bmp, tiff, heic), Videos (mp4, mov, avi), Audio (mp3, wav, aac, m4a, ogg, flac) Post production: AI music, TTS narration, voice cloning, lip sync, subtitles, transitions References Load these when needed: First time or config error → read references/SETUP CHECKLIST.md Error codes or failures → read references/TROUBLESHOOTING.md