MCP Usage Examples

This guide provides practical examples of using Browserman through the Model Context Protocol (MCP) with AI assistants.

Prerequisites

• MCP configured in your MCP client (Claude Code, Cursor, Codex, etc.)
• At least one account bound in Browserman
• Valid API key with appropriate permissions

Basic Examples

Posting to Twitter

Prompt to your AI assistant:

Post a tweet saying "Excited to announce our new product launch! 🚀"
using my twitter account named "jingo"

What the AI does:

1. Calls twitter_createTweet tool
2. Parameters: accountName: "jingo", text: "Excited to announce our new product launch! 🚀"
3. Returns the task ID and result

Engaging with Tweets

Reply to a Tweet:

Reply to this tweet https://x.com/username/status/123456
with "Great insight! Thanks for sharing."

Like a Tweet:

Like this tweet: https://x.com/username/status/123456

Retweet:

Retweet this: https://x.com/username/status/123456

Quote Tweet:

Quote tweet https://x.com/username/status/123456
with my comment "This is exactly what we've been working on!"

Multi-Platform Examples

Cross-Platform Posting

Prompt:

Post "Market analysis for today: bullish trends in tech sector"
to both my Twitter and Xueqiu accounts

Result: Your AI assistant will make multiple tool calls:

1. twitter_createTweet for Twitter
2. xueqiu_post for Xueqiu

Platform-Specific Content

Posting to Stock Platforms:

For Eastmoney (东方财富):

Post a comment on stock 600000 saying "Strong fundamentals,
good long-term hold" using my Eastmoney account

For Tonghuashun (同花顺):

Post to Tonghuashun: "Technical analysis shows bullish pattern"

Advanced Examples

Scheduled Posting Workflow

Prompt:

Help me create a content calendar. I want to post these messages:
- Monday: "Weekly market recap" to Twitter
- Wednesday: "Mid-week analysis" to Xueqiu
- Friday: "Weekend market outlook" to both platforms

Your AI assistant can help you plan and execute this workflow using the MCP tools.

Content Repurposing

Prompt:

Take this article [paste content] and create platform-specific
posts for Twitter (280 chars), Xueqiu (longer format), and
Eastmoney (stock-focused). Post to all platforms.

Your AI assistant will:

1. Reformat content for each platform
2. Call appropriate MCP tools for each platform
3. Track the results

Account Management

List All Accounts:

Show me all my connected accounts

List Platform-Specific Accounts:

Show me all my Twitter accounts

Check Task Status:

Check the status of task ID abc-123-def

Working with Images

Tonghuashun with Images

Prompt:

Post to Tonghuashun with text "Q4 earnings report analysis"
and these images:
https://example.com/chart1.jpg, https://example.com/chart2.jpg

Your AI assistant will use the tonghuashun_post tool with the urls parameter.

Error Handling Examples

Retry Failed Tasks

Prompt:

The last tweet failed to post. Can you retry with the same content?

Account Verification

Prompt:

Before posting, verify that my "jingo" Twitter account
is properly connected

Automation Patterns

Daily Engagement Routine

Prompt:

Create a daily engagement routine:
1. List my recent tweets
2. Engage with top replies
3. Post a morning update
4. Schedule an evening summary

Content Monitoring

Prompt:

Monitor these keywords on Twitter: "AI automation", "browser tools"
When you find relevant tweets, like and retweet the top 3

Best Practices

1. Be Specific with Account Names

Good:

Post to Twitter using my "personal" account

Bad:

Post to Twitter

2. Verify Before Bulk Actions

Good:

Show me all accounts first, then post to each Twitter account

Bad:

Post to all my accounts immediately

3. Handle Rate Limits

Good:

Post these 5 tweets with 30 seconds between each

Bad:

Post 50 tweets immediately

4. Use Full Mode When Needed

Good:

Post this sensitive content using full browser mode to avoid detection

Combining with Other Tools

With Web Search

Prompt:

Search for trending AI news, summarize the top 3 stories,
and post a thread to Twitter

With Content Generation

Prompt:

Generate 3 tweet variations about our new feature launch,
then post the best one to my Twitter account

With Data Analysis

Prompt:

Analyze my recent tweet engagement (provide data),
then create an optimized post for maximum engagement

Complex Workflows

Product Launch Campaign

Prompt:

Help me execute a product launch campaign:
1. Post teaser on Twitter
2. Share detailed analysis on Xueqiu
3. Post stock impact analysis on Eastmoney
4. Monitor and engage with responses
5. Post 24-hour summary

Multi-Account Management

Prompt:

I manage 5 Twitter accounts for different brands.
Create a posting schedule where each account posts
relevant content 3x per day without overlap.

Debugging Tips

Check Tool Availability

Prompt:

List all available Browserman MCP tools

Verify Configuration

Prompt:

Test my Browserman MCP connection and show me what
platforms I can access

Review Task History

Prompt:

Show me the status of my last 5 tasks

Example Conversations

Example 1: Brand Management

User: "I need to maintain our brand presence across platforms"

AI: "I can help! Let me first check what accounts you have connected."

[Calls accounts_list]

AI: "I see you have Twitter, Xueqiu, and Tonghuashun accounts. What would you like to post?"

User: "Post an update about our Q4 growth to all platforms"

[AI creates platform-specific content and posts to each]

Example 2: Crisis Response

User: "We need to respond to a situation quickly on Twitter"

AI: "I can help with that. What's the message?"

User: "We're aware of the issue and working on a fix. More updates soon."

AI: "Which account should I use for this urgent post?"

User: "Our official account 'company-official'"

[AI posts using full mode for reliability]

Rate Limits and Quotas

When using MCP tools, be aware of platform-specific rate limits:

• Twitter: ~300 tweets/3 hours
• Xueqiu: ~50 posts/day
• Eastmoney: ~100 comments/day
• Tonghuashun: ~30 posts/day

Your AI assistant can help you manage these limits by spacing out requests.

Next Steps

• Review API Documentation
• Troubleshoot Issues
• Learn About Account Binding