Use Draft Lift from AI coding agents (Claude Code, Cursor, VS Code) or your terminal.
Get up and running in 3 commands.
npm install -g @draftlift/cli
dl login
dl generate "Write a thread about AI productivity tips" --platform xOr set up MCP for your AI coding agent with one command:
dl setup --claude --mcpnpm install -g @draftlift/cliNo global install required — npx downloads it on demand.
npx @draftlift/mcpdl login # Opens browser for OAuth
dl login --no-browser # Prints URL to paste manuallyGenerate an API key in Settings and export it:
export DRAFTLIFT_API_KEY=dl_live_...Or pass it inline: dl generate --api-key dl_live_...
dl whoamiThe fastest way is dl setup, which auto-configures your agent:
dl setup --claude --mcp # Claude Code
dl setup --cursor # Cursor
dl setup --vscode # VS Code (Copilot)
dl setup --opencode # OpenCodeOr configure manually with the snippets below. Replace <your-api-key> with your API key.
claude mcp add draftlift \
--transport http \
--url https://api.draftliftai.com/mcp \
--header "Authorization: Bearer <your-api-key>"{
"mcpServers": {
"draftlift": {
"url": "https://api.draftliftai.com/mcp",
"headers": {
"Authorization": "Bearer <your-api-key>"
}
}
}
}{
"servers": {
"draftlift": {
"url": "https://api.draftliftai.com/mcp",
"headers": {
"Authorization": "Bearer <your-api-key>"
}
}
}
}{
"mcpServers": {
"draftlift": {
"url": "https://api.draftliftai.com/mcp",
"headers": {
"Authorization": "Bearer <your-api-key>"
}
}
}
}Run the MCP server as a local subprocess instead of connecting to the hosted endpoint:
{
"mcpServers": {
"draftlift": {
"command": "npx",
"args": ["-y", "@draftlift/mcp"],
"env": {
"DRAFTLIFT_API_KEY": "<your-api-key>"
}
}
}
}Run the MCP server locally with HTTP transport:
npx @draftlift/mcp --transport http --port 3100 --api-key dl_live_...Then point your agent to http://localhost:3100/mcp.
All commands follow the pattern dl <resource> <verb>. Top-level shortcuts exist for common actions.
Generate content for a platform.
dl generate "Write a thread about AI productivity" --platform x
dl generate "Launch announcement" --platform linkedin --template "Product Launch"
dl generate "..." --platform x --async
dl generate "..." --platform x --dry-run
dl generate "..." --platform x --jsonRepurpose existing content to a new platform or template.
dl repurpose <content-id> --platform linkedin
dl repurpose <content-id> --template "Blog Post" --dry-runList your content items.
dl content list
dl content list --platform x --status published --jsonGet a single content item by ID.
dl content get <content-id>
dl content get <content-id> --jsonScore content quality against platform rubrics.
dl score <content-id>
dl score <content-id> --jsonBrowse available templates.
dl templates list
dl templates list --platform x
dl templates list --jsonList all supported platforms.
dl platforms list
dl platforms list --jsonAdd a writing preference or brand voice memory.
dl memories add --platform x --content "Prefers casual tone"
dl memories listAdd reference material.
dl references add --url https://example.com/article
dl references add --text "Key talking points..." --title "Brand Guide"
dl references listManage workspaces.
dl workspace list
dl workspace use "My Agency"Schedule content for publication.
dl schedule <content-id> --date 2026-04-01 --time 09:00
dl schedule listCheck API usage and quotas.
dl usage
dl usage --jsonCheck job status (for async generation).
dl status <job-id>
dl status <job-id> --jsonConfigure Draft Lift for your AI coding agent.
dl setup --claude --mcp
dl setup --claude --cli
dl setup --cursor
dl setup --vscodeAuthenticate via browser OAuth.
dl login
dl login --no-browser
dl whoami
dl logout| Flag | Description |
|---|---|
| --json | Output as JSON (includes schema_version and trace_id) |
| --yes | Skip all confirmation prompts |
| --verbose | Show HTTP method, URL, status, and timing |
| --debug | Show full HTTP request/response headers and bodies |
| --dry-run | Preview what a command will do without executing (available on generate, repurpose, schedule, delete) |
| Code | Meaning |
|---|---|
| 0 | Success |
| 1 | User error (invalid flags, missing required args) |
| 2 | Auth error (expired key, not logged in) |
| 3 | Server error |
| 4 | Not found |