+ Post

Connect an agent

concivo has an MCP server, so your coding agent can build anyone's prompt in one line, and post the prompt behind the project it's working on, with screenshots, all from your terminal.

  1. 1

    Add concivo to your agent

    Claude Code: run this in a terminal
    claude mcp add --transport http --scope user concivo /api/mcp
    Other MCP clients (Cursor, Claude Desktop, …)
    {
  "mcpServers": {
    "concivo": {
      "type": "http",
      "url": "/api/mcp"
    }
  }
}
  2. 2

    Build or post

    To build a prompt, run claude "build handle/name from concivo" in an empty folder. It evolves it into your profile and builds it.

    To share yours, open the project and run /mcp__concivo__post, or just say “post this project to concivo”.

  3. 3

    Follow along

    The first time, your browser opens so you can sign in or create an account, and the agent helps you set up your profile. When posting, it reads the project and drafts the name, prompt and steps with you, asks for screenshots, and publishes only when you say so.

What the agent can do

  • start_building: build a prompt here in one go
  • start_posting: guided posting, step by step
  • create_prompt: save a private draft
  • get_upload_url, reorder_images: add and order screenshots and assets
  • publish_prompt, unpublish_prompt: make a draft public, or turn it back into a draft
  • update_prompt: edit fields, record revisions
  • get_prompt: read prompts you own or evolved
  • evolve_prompt: copy one into your profile
  • create_trait, list_traits: add-ons to pick when evolving
  • setup_github_sync: update the prompt on every push
  • list_my_prompts, whoami