manus

Verified·Scanned 2/17/2026

This skill integrates Manus AI into Clawdbot for autonomous research, development, automation, and media generation, providing scripts like run_task.py, upload_file.py, and webhook_server.py. It reads MANUS_API_KEY and runs python3 scripts that communicate with https://api.manus.ai/v1.

from clawhub.ai·v7a4aca6·29.2 KB·0 installs
Scanned from 0.1.0 at 7a4aca6 · Transparency log ↗
$ vett add clawhub.ai/disi3r/manus

Manus AI Skill for Clawdbot

🧠 Autonomous AI agent skill with web access, research, development, and automation capabilities

Overview

Manus AI is an autonomous AI agent capable of executing complex tasks, researching, developing software, automating workflows, and generating multimedia content.

This skill provides full integration with Manus API for Clawdbot/OpenClaw agents.

Features

  • 🧠 Autonomous Tasks - Complex task execution with progress tracking
  • 🔍 Research - Web search, data analysis, and information synthesis
  • 💻 Development - Build websites, apps, and software solutions
  • 🎨 Media Generation - Images, videos, audio from text descriptions
  • 🔄 Workflow Automation - Bookings, data collection, scheduled tasks
  • 📊 Data Integrations - Similarweb, premium data sources
  • 🔗 Connectors - Gmail, Notion, Google Calendar, Slack

Installation

Via Clawdbot (recommended)

clawdbot skills install manus

Manual Installation

# Clone the repository
git clone https://github.com/disier/clawdbot-skill-manus.git
cd clawdbot-skill-manus

# Install dependencies
npm install

Configuration

API Key

Set your Manus API key:

export MANUS_API_KEY="sk-..."

Or in ~/.clawdbot/clawdbot.json:

{
  "skills": {
    "manus": {
      "apiKey": "sk-..."
    }
  }
}

Get your API key at: manus.im/app

Usage

Basic Task

cd scripts
python3 run_task.py "Research the latest AI regulations in the EU"

With Connectors

# Gmail
python3 run_task.py "Read my recent emails and summarize"

# Notion
python3 run_task.py "Create a project page in Notion"

# Google Calendar
python3 run_task.py "Schedule a meeting for tomorrow at 3pm"

# Slack
python3 run_task.py "Post an update to #announcements"

File Upload

# Upload context file
python3 upload_file.py context.md

# Use with task
python3 run_task.py "Analyze this data and create a report"

Check Status

python3 check_status.py TASK_ID

Get Results

python3 get_result.py TASK_ID

Webhook Server

# Start webhook server on port 8080
python3 webhook_server.py 8080

# Register webhook with Manus
python3 webhook_server.py 8080 --register --url https://your-domain.com/webhook/manus

Scripts

ScriptDescription
run_task.pyExecute a task with progress tracking
create_project.pyCreate a project
upload_file.pyUpload files for context
check_status.pyCheck task status
get_result.pyGet task result
webhook_server.pyReal-time notifications server

Integration with Clawdbot

In an Agent

When deep research or development is needed:
1. Use the run_task.py script from manus skill
2. Provide clear prompt
3. Wait for completion
4. Integrate results

Example

# Research and create content
python3 run_task.py "Research 5 tech trends for 2026 and write a 1000-word article"

# With context
python3 upload_file.py data.csv
python3 run_task.py "Analyze this CSV and generate a sales report"

OpenAI Compatibility

Manus is compatible with the OpenAI SDK:

from openai import OpenAI

client = OpenAI(
    api_key="sk-...",  # Your Manus API key
    base_url="https://api.manus.ai/v1"
)

response = client.chat.completions.create(
    model="manus-1.6-adaptive",
    messages=[{"role": "user", "content": "Hello!"}]
)

API Reference

Endpoints

EndpointDescription
POST /v1/tasksCreate a new task
GET /v1/tasks/{id}Get task status/result
POST /v1/projectsCreate a project
POST /v1/filesUpload files
POST /v1/webhooksRegister webhooks

Models

  • manus-1.6-adaptive (default)
  • manus-1.5-pro

Best Practices

Effective Prompts

✅ Good:

  • "Research AI regulations in the EU and summarize key points"
  • "Create a weather web app with React and OpenWeatherMap"
  • "Analyze the last 10 tweets from an account"

❌ Avoid:

  • "Do something useful" (too vague)
  • "Improve this" (no context)

Large Tasks

For long-running tasks:

python3 run_task.py "Deep research on AI market" --timeout 300

Contributing

  1. Fork the repository
  2. Create a feature branch
  3. Commit your changes
  4. Push to the branch
  5. Open a Pull Request

License

MIT License - see LICENSE for details.

Support


Built with ❤️ by DisierTECH