VisionStory Docs
llms.txt Get API key

For agents

View as MarkdownOpen in ClaudeOpen in ChatGPT

AI agents can drive the VisionStory API through three channels. All three use the same API key and the same REST API underneath — pick the one that fits where your agent runs.

Choose a channel

Channel What it is Best when Status
llms.txt + .md pages Machine-readable docs index; every docs page is also served as Markdown Your agent browses the web and reads docs on demand Available
Agent Skill Installable package: SKILL.md instructions + a dependency-free Python CLI Coding agents (Claude Code, Codex, …) working inside a repo Available
MCP server Task-oriented tools (generate_video, create_avatar, …) over the Model Context Protocol Chat agents like Claude Desktop, no code execution needed Available

Whichever channel you use, configure the key the same way — the agent reads it from an environment variable and never pastes it into chat:

export VISIONSTORY_API_KEY="sk-vs-xxxxxxxxxxxxxxxxxxx"

llms.txt

Point any web-capable agent at the docs index; it links every page in agent-readable form:

Read https://openapi.visionstory.ai/docs/llms.txt for the VisionStory API docs index.
Append .md to any guide or endpoint page URL to get that page as Markdown.
Use https://openapi.visionstory.ai/docs/llms-full.txt for the entire docs in one file.

You can also hand a single page to an agent directly — every page has a "Copy page" menu with Copy as Markdown, Open in Claude, and Open in ChatGPT.

Agent Skill

Install the skill package into your project — three commands:

mkdir -p .agents/skills
curl -fsSL https://openapi.visionstory.ai/skills/visionstory-video-api.zip -o visionstory-video-api.zip
unzip -o visionstory-video-api.zip -d .agents/skills && rm visionstory-video-api.zip

The package contains SKILL.md (workflow instructions the agent follows) and scripts/visionstory_api.py, a zero-dependency Python CLI with consistent auth, base64 encoding, polling, timeouts, and downloads. Skills-compatible agents pick it up automatically; you can also view SKILL.md without downloading.

The CLI works standalone too:

python3 .agents/skills/visionstory-video-api/scripts/visionstory_api.py models
python3 .agents/skills/visionstory-video-api/scripts/visionstory_api.py create-video --avatar-id 4321918387609092991 --text "Hello from VisionStory." --voice-id Alice --output result.mp4

MCP server

visionstory-mcp exposes task-oriented tools — generate_video (submit + poll in one call), get_video, create_avatar, list_avatars, clone_voice, list_voices, list_models, get_credits, and upload_asset — so chat agents can produce videos without writing code. Add it to Claude Desktop (or any MCP client):

{
  "mcpServers": {
    "visionstory": {
      "command": "uvx",
      "args": ["visionstory-mcp"],
      "env": { "VISIONSTORY_API_KEY": "sk-vs-xxxxxxxxxxxxxxxxxxx" }
    }
  }
}

The visionstory-mcp package is being published to PyPI. Until it lands, run the server from the repository with python -m src.open_api.mcp (same VISIONSTORY_API_KEY environment variable).

The server reads the key from VISIONSTORY_API_KEY; generate_video blocks until the task finishes and returns the final video URL, so the agent gets one call in, one result out.

End-to-end example

With the Skill installed (or the MCP server connected), a single instruction produces a video:

Create a talking-avatar video that says "Welcome to our launch week!", pick a friendly public avatar and an energetic voice, wait for it to finish, and save it as launch.mp4.

The agent will list avatars and voices, submit POST /api/v1/video, poll GET /api/v1/video?video_id=... until created, and download the video_url — the same flow as the Quick start.

Rules your agent should follow

Next steps