TokPortal MCP Server: Let AI Agents Manage Your TikTok Distribution

The Model Context Protocol (MCP) is an open standard that lets AI assistants like Claude, ChatGPT, and custom agents interact with external tools and services in real time. Instead of just generating text about what you could do, an MCP-connected AI agent can actually take actions on your behalf — creating accounts, uploading videos, checking analytics, and managing entire campaigns through natural conversation.

TokPortal's MCP server exposes the full power of the TokPortal API to any MCP-compatible AI client. This means you can tell Claude to "create 5 TikTok accounts in France with fitness-themed profiles" and it will actually execute those actions — calling the API, handling responses, and reporting back with real results. No dashboards, no manual API calls, no scripting required.

This represents a fundamental shift in how social media campaigns are managed. Instead of learning API documentation or building complex automation workflows, you simply describe what you want in plain English and let the AI agent handle the technical execution. It's the most accessible way to leverage TokPortal's capabilities, and it's available today.

Setting Up TokPortal MCP

claude_desktop_config.json

// claude_desktop_config.json
// Location: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
// Location: %APPDATA%\Claude\claude_desktop_config.json (Windows)

{
  "mcpServers": {
    "tokportal": {
      "command": "npx",
      "args": ["-y", "@tokportal/mcp-server"],
      "env": {
        "TOKPORTAL_API_KEY": "your_api_key_here"
      }
    }
  }
}

What AI Agents Can Do

Create Accounts Instantly — Tell the agent to create TikTok or Instagram accounts in any supported country. It handles bundle creation, country selection, and platform configuration through simple conversational commands.
Configure Profiles — Describe the persona you want — username, display name, bio, profile picture — and the agent configures everything via the API. It can even suggest usernames based on your niche.
Upload & Schedule Videos — Share video URLs or file paths, and the agent uploads them to TokPortal, sets captions and hashtags, and schedules them for optimal posting times across your accounts.
Schedule Posts Strategically — Ask the agent to spread posts across accounts with optimal timing. It understands time zones, peak engagement hours, and can stagger uploads to avoid detection patterns.
Check Analytics in Real Time — Request performance reports in natural language — the agent pulls follower counts, engagement rates, view metrics, and video-level analytics, then summarizes the data with actionable insights.
Manage Account Warming — The agent can enable niche warming with custom instructions, check warming status, and recommend deep warming for Instagram accounts that need stronger engagement signals.
Handle Video Corrections — When videos need edits or re-uploads, the agent can request corrections, provide feedback to account managers, and re-submit fixed content — all through conversation.
Monitor Campaign Health — Ask for a campaign overview and the agent checks all your bundles, flags any issues (broken download links, pending reviews), and suggests next steps to keep everything on track.

Example: Autonomous Campaign Management

The real power of MCP becomes clear when you see an AI agent managing a complete campaign from start to finish. Instead of navigating dashboards or writing scripts, you have a natural conversation where the agent takes real actions at each step. Below is an example of how a conversation with Claude might look when using the TokPortal MCP server.

Notice how the agent doesn't just suggest what to do — it actually executes each step, confirms the results, and moves on to the next action. The entire campaign setup that would normally take an hour of manual work happens in a few minutes of conversation. The agent handles API calls, error checking, and status verification behind the scenes.

claude-mcp-conversation.txt

User: Create 3 TikTok accounts in Germany for a fitness brand.
       Use niche warming focused on gym and workout content.

Claude: I will create 3 TikTok account bundles in Germany with
        fitness-focused niche warming. Let me set those up now.

        [Calling create_bundle × 3 with platform: tiktok,
         country: DE, wants_niche_warming: true]

        Done! I have created 3 bundles:
        • Bundle 1: abc-123 (pending_setup)
        • Bundle 2: def-456 (pending_setup)
        • Bundle 3: ghi-789 (pending_setup)

        Each has niche warming enabled for gym/workout content.
        Would you like me to configure the account profiles now?

User: Yes, make them fitness-themed with German usernames.

Claude: I will configure each account with fitness-themed profiles.

        [Calling configure_account × 3]

        All three accounts are configured:
        • fit.life.berlin — "Berlin Fitness" — Gym tips & Workouts
        • daily.workout.de — "Tägliches Training" — Home workouts
        • gym.motivation.de — "Gym Motivation" — Fitness inspo

        Ready to publish these bundles?

User: Yes, publish them all.

Claude: [Calling publish_bundle × 3]

        All 3 bundles are now published and assigned to account
        managers. You will receive the accounts within 24-48 hours.
        I can check back on their status anytime — just ask.

Safety first: monitor your AI agents

While AI agents are powerful, they should always operate with human oversight. Review the actions your agent takes, especially for operations that spend credits like creating bundles or adding video slots. Start with small campaigns to build confidence in the agent's decisions before scaling up.

Beyond Claude: Custom AI Agents

While Claude Desktop is the most popular MCP client, the TokPortal MCP server works with any MCP-compatible application. The protocol is open and standardized, which means developers can build custom AI agents that leverage TokPortal's capabilities in any programming language or framework.

For example, you could build a Python-based agent using the MCP SDK that runs on a schedule — checking campaign performance every morning, creating new accounts when engagement metrics hit certain thresholds, and sending weekly reports to your team via email. The agent operates autonomously within the guardrails you define, using the same TokPortal MCP tools that Claude uses.

Enterprise teams are already building internal AI assistants that combine TokPortal MCP with other business tools — CRM data for client management, analytics platforms for cross-channel reporting, and content management systems for asset libraries. The MCP standard makes it possible to give a single AI agent access to your entire marketing stack, with TokPortal handling the social media distribution layer.

Give your AI agent superpowers

Connect TokPortal to Claude, ChatGPT, or any MCP-compatible AI agent. Manage your entire TikTok and Instagram distribution through natural conversation.

Explore the MCP Server

What is the Model Context Protocol (MCP)?+

MCP is an open standard developed by Anthropic that allows AI assistants to interact with external tools and services. Instead of just generating text, an MCP-connected AI can take real actions — like calling APIs, reading databases, or managing files. TokPortal's MCP server exposes its API as MCP tools, letting AI agents create accounts, post videos, and manage campaigns.

Which AI clients support TokPortal MCP?+

Any MCP-compatible client works with TokPortal's MCP server. Currently, the most popular options are Claude Desktop (by Anthropic), Cursor IDE, and custom agents built with the MCP SDK in Python, TypeScript, or other languages. As the MCP ecosystem grows, more AI clients will be compatible.

Is it safe to let AI agents manage my TokPortal account?+

The MCP server uses your API key for authentication, so the agent has the same permissions you do. We recommend starting with read-only operations (checking analytics, listing bundles) before moving to write operations (creating accounts, publishing videos). You can also set spending limits in your TokPortal dashboard to prevent unexpected credit usage.

Can AI agents handle complex multi-step campaigns?+

Yes. MCP-connected agents can chain multiple TokPortal operations together — creating bundles, configuring accounts, uploading videos, scheduling posts, and monitoring analytics — all in a single conversation. The agent maintains context across steps, so it can reference previously created bundles and build on earlier actions.

Written by

Vincent Tellenne

Founder & CEO

Vincent is the founder of TokPortal, building the infrastructure for scaled organic social media distribution. Previously scaled multiple startups and APIs to millions of requests.

@tellenne_LinkedIn

Learn more about this topic with AI

Ask ChatGPT Ask Perplexity Ask Claude