Liftli MCP Server
Acts as a head of content for LinkedIn, X (Twitter), and Substack, enabling voice extraction, content ideation, drafting, critiquing, and approved publishing/scheduling through official APIs.
README
Liftli — head of content, as an MCP server
<p align="center"> <img alt="LinkedIn" src="https://cdn.worldvectorlogo.com/logos/linkedin-icon-2.svg" width="32"> <img alt="X (Twitter)" src="https://cdn.simpleicons.org/x/000000" width="32"> <img alt="Substack" src="https://cdn.simpleicons.org/substack/FF6719" width="32"> </p>
<p align="center"><strong>LinkedIn · X (Twitter) · Substack</strong></p>
Liftli is a remote MCP server that acts as a head of content for LinkedIn, X (Twitter) and Substack, inside the AI assistant you already use.
It extracts your writing voice from your own posts, mines your real work — voice notes, meeting transcripts, GitHub activity, chats, the news — for post ideas, builds and remembers your strategy, drafts in your voice, critiques every draft like a ruthless editor, and schedules or publishes through the platforms' official APIs.
Nothing publishes without your explicit approval. No scraping, no browser automation, no bots acting on your profile.
- Homepage: https://liftli.ai
- Developer docs: https://liftli.ai/developers.html
- Free skills (no account needed): https://github.com/liftli-ai/skills
About this repository. Liftli is a hosted service; this repo is documentation and install instructions for the remote MCP server. The server implementation is not open source. If you are looking for something self-hostable, this is not it — but the connector is free to add and starts with a free usage allowance, no card required.
Install
The server is remote — served over HTTPS, using MCP's Streamable HTTP transport. There is nothing to run locally.
Connector URL
https://liftli-strategist-mcp.onrender.com/mcp
Claude Code
claude mcp add --scope user --transport http liftli https://liftli-strategist-mcp.onrender.com/mcp
Claude Desktop — Customize (briefcase icon) → Connectors → + Add custom connector → paste the URL above into "Remote MCP Server URL".
Other clients (ChatGPT, Cursor, Codex, Copilot, Gemini) — add it as a custom/remote MCP server using the same URL. Client-specific notes are in developers.html.
If your client cannot reach a remote server directly, npx mcp-remote works as a bridge.
Tools
54 tools. Some clients expose a reduced set (ChatGPT 53, Cursor 40) to stay within their limits.
Start here / profile (7) — strategist_briefing, get_profile, save_profile, onboard,
extract_voice, save_corpus_tags, liftli_status
Strategy (6) — build_strategy, save_strategy, define_narrative, growth_strategy,
plan_calendar, content_ideas
Writing (7) — plan_post, generate_hooks, critique, ready_to_post, save_post,
browse_posts, rewrite_profile
Engagement (4) — plan_comment, engagement_plan, analyze_results, log_outcome
Substack (9) — build_substack_strategy, save_substack_strategy, plan_newsletter,
score_newsletter, newsletter_headlines, plan_note, notes_from_newsletter,
repurpose_to_substack, substack_growth
Publishing & scheduling (7) — connect_account, approve_post, publish_post,
schedule_post, list_scheduled, edit_scheduled, send_post_to_telegram
Media & visuals (3) — generate_visual, plan_visual, manage_media
Memory & seeds (5) — remember, recall, manage_memory, extract_seeds, news_monitor
Channels & account (6) — link_telegram, link_whatsapp, notify, contact_support,
upgrade, dismiss_suggestion
How it differs
Liftli covers the whole pipeline: it decides what is worth saying from your real week, drafts it in your extracted voice, pressure-tests the draft, and then publishes or schedules it through the platforms' official APIs. You do not need to add a separate scheduler.
The dividing line against other tools is where it lives, not what it does. Schedulers and dashboards are surfaces you open and sit inside. Liftli is not a surface — it runs inside the AI assistant you already use, so capturing an idea, writing it and shipping it happen in one place, including from a voice note sent over Telegram or WhatsApp while you are away from a desk.
Safety model
- Publishing and scheduling run through official platform APIs, connected only when you explicitly link an account.
- Every draft waits for your approval before anything is posted.
- No browser extension, no scraping, no automated actions on your profile — the category the platforms' terms prohibit.
Support
Open an issue here, or use the contact_support tool from inside your assistant.
License
Documentation in this repository is MIT-licensed. The Liftli service itself is proprietary and governed by the terms at liftli.ai.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。