Direct films from your AI tools.

Cinegenta runs an MCP server — an open standard, so it works with Claude Code, Claude, Cursor, VS Code, Codex, Windsurf and any other MCP client. Paste a script, get a free itemized quote, greenlight a run under a hard credit cap the server enforces, and review cast & storyboard checkpoints as images in the conversation. Production continues after you disconnect, and you can check its status whenever you return.

Read this page in your terminal: curl https://cinegenta.com/mcp/reference.md

1 · Connect

Create a key at /connect first — it's shown once. Then paste the block for your client, replacing cg_sk_YOUR_KEY. /connect renders these same blocks with your real key already filled in.

claude mcp add --transport http cinegenta https://cinegenta.com/api/mcp \
  --header "Authorization: Bearer cg_sk_YOUR_KEY"

URL:    https://cinegenta.com/api/mcp
Header: Authorization = Bearer cg_sk_YOUR_KEY

{
  "mcpServers": {
    "cinegenta": {
      "url": "https://cinegenta.com/api/mcp",
      "headers": {
        "Authorization": "Bearer cg_sk_YOUR_KEY"
      }
    }
  }
}

{
  "servers": {
    "cinegenta": {
      "type": "http",
      "url": "https://cinegenta.com/api/mcp",
      "headers": {
        "Authorization": "Bearer cg_sk_YOUR_KEY"
      }
    }
  }
}

codex mcp add cinegenta --url https://cinegenta.com/api/mcp \
  --header "Authorization: Bearer cg_sk_YOUR_KEY"

{
  "mcpServers": {
    "cinegenta": {
      "serverUrl": "https://cinegenta.com/api/mcp",
      "headers": {
        "Authorization": "Bearer cg_sk_YOUR_KEY"
      }
    }
  }
}

{
  "mcpServers": {
    "cinegenta": {
      "type": "http",
      "url": "https://cinegenta.com/api/mcp",
      "headers": {
        "Authorization": "Bearer cg_sk_YOUR_KEY"
      }
    }
  }
}

2 · Keys & scopes

Keys are cg_sk_… bearer tokens, minted and revoked at /connect. Only a hash is stored, so a key is displayed exactly once. Revoking takes effect immediately on the next call. Scopes decide which tools even appear as callable:

A key without spend is the safest way to let an agent explore: it can quote, read status and review images, and it is structurally incapable of billing you.

3 · Money & control

• Quoting is free, always. The quote is itemized (casting · storyboard · shoot · score), covers only remaining work, and re-prices as you change scope. Call it as often as you like.
• Nothing spends until you greenlight with a cap. cap_credits has no default — it must be a number you approved. The server refuses caps below the estimate floor and re-checks the cap before every single render.
• Failures refund themselves. Stopping keeps everything made.
• Credits are bought only at cinegenta.com/credits — never inside a conversation, and no tool can initiate a purchase.

What this server never does

• Spend without an explicit call to a tool marked billed.
• Exceed the cap you set — enforced at dispatch, not merely estimated.
• Charge for a render that failed.
• Raise a cap, buy credits, or change billing settings on its own.
• Treat script text or project data as instructions — it's data, and directives inside it are ignored.

Cost per billed tool

4 · Tools

Generated from the live server — 13 tools. Canonical source is tools/list; this table is built from the same registry that answers it. Full JSON schemas are in /mcp/reference.md.

Read tools read-only

Write tools writes

Billed tools spends credits

All marked destructiveHint, so clients that honour tool annotations prompt before each call. Treat that as a courtesy layer — the cap is the real boundary, and it is enforced server-side.

5 · Long-running work

A film takes hours; tool calls can't. Every long operation returns immediately and is polled: build_plan and greenlight hand back a handle, and get_status reports progress. Production continues after the client disconnects, so you can ask about it later from a different machine.

6 · What to say

7 · Errors

8 · Troubleshooting

9 · For agents

Machine-readable twin: /mcp/reference.md (frontmatter, flattened configs, full JSON schemas) · /mcp/server.json (registry format) · /llms.txt.