Build AI agents and modern applications with the platform for the AI era, fully integrated with your agentic coding tools.
Functions, frontends, databases, and the rest of the platform are documented under Platform.
The same docs are in your terminal with volcano docs.
For step-by-step recipes you can build, explore the Cookbooks.
For application design ideas and practical perspectives, visit the Blog.
Install the CLI, sign up, create a project, and start building.
Deploy and manage functions, frontends, and databases from your terminal or CI.
Do everything the dashboard does over HTTP, with an OpenAPI spec to generate clients.
Connect your coding agent to a project to read its deployments and logs, and change it.
Install Volcano skills and plugins for your coding agent.
Your coding agent can inspect a project's functions, deployments, and logs, and change it with a full-scope token. Create a project access token, export it, and add the server to your client.
export VOLCANO_PROJECT_TOKEN=pt-your-project-access-tokenStart your agent from that shell. Editors such as Cursor and VS Code read the variable from the environment they were launched in, so quit the editor fully and reopen it from the shell, for example with cursor . or code .. Exporting it in the editor's own terminal is not enough.
claude mcp add --transport http volcano https://api.volcano.dev/mcp \
--header 'Authorization: Bearer ${VOLCANO_PROJECT_TOKEN}'Project-level config files, token scopes, and the full tool list are in the MCP server guide.