Skip to main content
Glama
nanameru

Manus MCP

by nanameru
OverviewSchemaRelated ServersScoreDiscussions
JavaScript
Remote

Manus MCP

The Universal MCP Server for Manus AI enables you to create AI tasks, manage webhooks, and integrate Manus workflows into any MCP-compatible client. Designed for prompt-first usage with full support for attachments, connectors, and real-time notifications.

Installation

Prerequisites

  • Node.js 18+

  • Set MANUS_MCP_API_KEY in your environment

Get an API key

Build locally

cd /path/to/manus-mcp
npm i
npm run build

Related MCP server: Anam MCP Server

Setup: Claude Code (CLI)

Use this one-liner (replace with your real API key):

claude mcp add manus-mcp -s user -e MANUS_MCP_API_KEY="your-api-key-here" -- npx manus-mcp

Note: Use manus-mcp (not "Manus MCP") as the name. Claude CLI requires names without spaces.

To remove:

claude mcp remove manus-mcp

Setup: Cursor

Create .cursor/mcp.json in your client (do not commit it here):

{
  "mcpServers": {
    "manus-mcp": {
      "command": "npx",
      "args": ["manus-mcp"],
      "env": { "MANUS_MCP_API_KEY": "your-api-key-here" },
      "autoStart": true
    }
  }
}

Note: This repository does not include .cursor/mcp.json. Configure Cursor via the UI or manually create the file in your client workspace.

Other Clients and Agents

Install via URI or CLI:

code --add-mcp '{"name":"manus-mcp","command":"npx","args":["manus-mcp"],"env":{"MANUS_MCP_API_KEY":"your-api-key-here"}}'

Or add to your VS Code settings JSON:

{
  "mcp.servers": {
    "manus-mcp": {
      "command": "npx",
      "args": ["manus-mcp"],
      "env": { "MANUS_MCP_API_KEY": "your-api-key-here" }
    }
  }
}

Same as VS Code, but use code-insiders command:

code-insiders --add-mcp '{"name":"manus-mcp","command":"npx","args":["manus-mcp"],"env":{"MANUS_MCP_API_KEY":"your-api-key-here"}}'

Add to your Claude Desktop configuration file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

  • Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "manus-mcp": {
      "command": "npx",
      "args": ["manus-mcp"],
      "env": { "MANUS_MCP_API_KEY": "your-api-key-here" }
    }
  }
}

In LM Studio's MCP settings:

  • Command: npx

  • Args: ["manus-mcp"]

  • Env: MANUS_MCP_API_KEY=your-api-key-here

Add to your Goose configuration:

  • Type: STDIO

  • Command: npx

  • Args: manus-mcp

  • Enabled: true

  • Env: MANUS_MCP_API_KEY=your-api-key-here

Example ~/.config/opencode/opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "manus-mcp": {
      "type": "local",
      "command": ["npx", "manus-mcp"],
      "enabled": true,
      "env": { "MANUS_MCP_API_KEY": "your-api-key-here" }
    }
  }
}

Add a new MCP in Qodo Gen and paste the standard JSON config:

{
  "command": "npx",
  "args": ["manus-mcp"],
  "env": { "MANUS_MCP_API_KEY": "your-api-key-here" }
}

Follow the Windsurf MCP integration guide and reuse the standard config:

{
  "manus-mcp": {
    "command": "npx",
    "args": ["manus-mcp"],
    "env": { "MANUS_MCP_API_KEY": "your-api-key-here" }
  }
}

Setup: Codex (TOML)

Add the following to your Codex TOML configuration:

[mcp_servers.manus-mcp]
command = "npx"
args = ["manus-mcp"]

[mcp_servers.manus-mcp.env]
MANUS_MCP_API_KEY = "your-api-key-here"
# MCP_NAME = "manus-mcp"  # Optional: override server name

Configuration (Env)

  • MANUS_MCP_API_KEY: Your Manus API key (required)

  • MANUS_MCP_API_BASE_URL: Override the API base URL (default: https://api.manus.ai/v1)

  • MCP_NAME: Server name override (default: manus-mcp)

Available Tools

create_task

Create a new AI task in Manus with custom parameters and optional attachments.

Inputs:

{
  "prompt": "string (required) - The task prompt or instruction for the AI",
  "mode": "string (required) - 'speed' or 'quality'",
  "attachments": "array (optional) - List of attachment objects { filename, url, mime_type, size_bytes }",
  "connectors": "array (optional) - List of connector IDs (e.g., ['gmail', 'notion'])",
  "hide_in_task_list": "boolean (optional) - Hide from webapp task list (default: false)",
  "create_shareable_link": "boolean (optional) - Generate a public shareable link (default: false)"
}

Outputs:

{
  "task_id": "string - Unique task identifier",
  "task_title": "string - Generated task title",
  "task_url": "string - Direct link to the task",
  "shareURL": "string (optional) - Public shareable link if requested"
}

create_webhook

Register a new webhook to receive real-time notifications from Manus.

Inputs:

{
  "url": "string (required) - Webhook endpoint URL",
  "events": "array (optional) - List of event types to subscribe to"
}

Outputs: Webhook registration details including webhook ID and configuration.

delete_webhook

Remove a previously registered webhook by its ID.

Inputs:

{
  "webhook_id": "string (required) - The ID of the webhook to delete"
}

Outputs: Success confirmation or error details.

Example invocation (MCP tool call)

Create a task in speed mode:

{
  "tool": "create_task",
  "arguments": {
    "prompt": "Analyze the quarterly sales data and generate a summary report with key insights",
    "mode": "speed",
    "create_shareable_link": true
  }
}

Create a task with attachments:

{
  "tool": "create_task",
  "arguments": {
    "prompt": "Extract key action items from this meeting transcript",
    "mode": "quality",
    "attachments": [
      {
        "filename": "meeting-notes.pdf",
        "url": "https://example.com/files/meeting-notes.pdf",
        "mime_type": "application/pdf",
        "size_bytes": 245632
      }
    ],
    "connectors": ["gmail", "notion"]
  }
}

Troubleshooting

401 Authentication Error

  • Verify that MANUS_MCP_API_KEY is correctly set in your environment

  • Check that your API key is valid and has not expired

  • Test your API key with a direct curl request:

    curl -H "API_KEY: your-key" https://api.manus.ai/v1/tasks

Node.js Version Error

  • Ensure you're using Node.js 18 or later: node -v

  • Update Node.js if necessary: https://nodejs.org/

Build Issues

  • Clear the build directory: rm -rf build

  • Reinstall dependencies: rm -rf node_modules && npm i

  • Rebuild: npm run build

Testing Local Builds

  • After building, test the server: npx . or node build/index.js

  • Check that the executable is properly created: ls -la build/index.js

Inspecting Publish Artifacts

  • See what would be published: npm pack --dry-run

  • Check the package contents: npm pack && tar -xzf manus-mcp-*.tgz && cat package/package.json

References

Name Consistency & Troubleshooting

Always use CANONICAL_ID (manus-mcp) for identifiers and keys. Use CANONICAL_DISPLAY (Manus MCP) only for UI labels. Do not mix legacy keys after registration.

Consistency Matrix

Context

Value

npm package name

manus-mcp

Binary name

manus-mcp

MCP server name (SDK metadata)

manus-mcp

Env default MCP_NAME

manus-mcp

Client registry key

manus-mcp

UI label

Manus MCP

Conflict Cleanup

  • Remove any stale keys (e.g., old display names like "Manus") and re-add with manus-mcp only

  • Ensure global .mcp.json or client registries only use manus-mcp for keys

  • Cursor: Configure in the UI; this project intentionally omits .cursor/mcp.json

Example

✅ Correct:

{
  "mcpServers": {
    "manus-mcp": {
      "command": "npx",
      "args": ["manus-mcp"]
    }
  }
}

❌ Incorrect:

{
  "mcpServers": {
    "Manus": {  // Wrong: will conflict with "manus-mcp"
      "command": "npx",
      "args": ["manus-mcp"]
    }
  }
}

License

MIT

Available Tools

3 tools
create_taskB

Create a new AI task in Manus. Returns task_id, task_title, task_url, and optionally a shareable link.

ParametersJSON Schema
NameRequiredDescriptionDefault
promptYesThe task prompt or instruction for the AI
modeYesExecution mode: 'speed' for faster results, 'quality' for better accuracy
attachmentsNoOptional attachments (files, URLs, etc.)
connectorsNoList of connector IDs to enable for this task (only pre-configured connectors)
hide_in_task_listNoWhether to hide this task from the Manus webapp task list (default: false)
create_shareable_linkNoWhether to make the chat publicly accessible (default: false)

TDQS

B3.1/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions that it 'Returns task_id, task_title, task_url, and optionally a shareable link,' which gives some output context, but lacks details on permissions, rate limits, side effects, or error handling for a creation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that front-loads the core action and key return values, with no wasted words. It effectively communicates the essential information in a compact form.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity of a creation tool with 6 parameters and no annotations or output schema, the description is minimally adequate. It covers the basic purpose and return values but lacks behavioral context and usage guidelines, leaving gaps for an AI agent to infer details.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, so the schema fully documents all 6 parameters. The description adds no additional meaning beyond the schema, such as explaining interactions between parameters or usage examples, meeting the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Create a new AI task in Manus') and specifies the resource ('AI task'), which is distinct from sibling tools like create_webhook and delete_webhook. However, it doesn't explicitly differentiate from siblings beyond the resource type, missing a direct comparison.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives, such as when to create a task versus a webhook, or any prerequisites like authentication needs. It only mentions the return values without context for usage decisions.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

create_webhookC

Register a new webhook to receive real-time notifications from Manus. Returns webhook details.

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesThe webhook URL endpoint to receive notifications
eventsNoList of event types to subscribe to

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It states the tool registers a webhook and returns details, but lacks critical information such as authentication requirements, rate limits, whether the registration is persistent, or error handling. This is insufficient for a mutation tool with zero annotation coverage.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that front-loads the key action and outcome with zero waste. It directly states what the tool does and the result, making it appropriately sized and well-structured.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity as a mutation tool with no annotations and no output schema, the description is incomplete. It lacks details on behavioral traits, error conditions, or what the returned 'webhook details' include, leaving significant gaps for an AI agent to understand and use the tool effectively.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, documenting both parameters ('url' and 'events') clearly. The description does not add any additional meaning or context beyond what the schema provides, such as examples or constraints, so it meets the baseline score of 3 for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: 'Register a new webhook to receive real-time notifications from Manus.' It specifies the verb ('register'), resource ('webhook'), and outcome ('receive real-time notifications'), but does not explicitly differentiate it from sibling tools like 'delete_webhook' beyond the action verb.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives. It mentions the outcome but does not specify prerequisites, context, or exclusions, such as when to choose this over other notification methods or how it relates to sibling tools like 'create_task'.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

delete_webhookC

Remove a previously registered webhook by its ID.

ParametersJSON Schema
NameRequiredDescriptionDefault
webhook_idYesThe ID of the webhook to delete

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries full burden for behavioral disclosure. It states this is a removal/deletion operation, implying it's destructive, but doesn't clarify whether deletion is permanent, reversible, or has side effects. No information about permissions, rate limits, or response format is included, leaving significant gaps for a mutation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that communicates the core purpose without unnecessary words. It's appropriately sized for a simple deletion tool and front-loads the essential information. Every word earns its place with no redundancy or fluff.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a destructive operation with no annotations and no output schema, the description is insufficient. It doesn't explain what happens after deletion, whether confirmation is required, or what the response contains. Given the tool's potential impact and lack of structured metadata, more behavioral context is needed to make it complete for safe agent use.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the parameter 'webhook_id' is fully documented in the schema. The description adds no additional semantic context about the parameter beyond what the schema provides ('The ID of the webhook to delete'). This meets the baseline expectation when schema coverage is complete.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Remove') and target ('previously registered webhook by its ID'), making the purpose immediately understandable. It doesn't explicitly differentiate from sibling tools like 'create_webhook', but the verb 'Remove' versus 'create' provides implicit distinction. The description avoids tautology by specifying what gets removed rather than just restating the tool name.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives or what prerequisites exist. While 'previously registered' implies the webhook must exist, it doesn't specify conditions for deletion or warn about consequences. There's no mention of sibling tools like 'create_webhook' for comparison or when deletion might be appropriate versus modification.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

B3.2/5.0
Disambiguation5/5

Each tool has a clearly distinct purpose: create_task handles AI task creation, create_webhook manages webhook registration, and delete_webhook handles webhook removal. There is no overlap in functionality, making tool selection straightforward for an agent.

Naming Consistency5/5

All tools follow a consistent verb_noun pattern (create_task, create_webhook, delete_webhook) with clear, descriptive names. The naming convention is uniform throughout the set, enhancing readability and predictability.

Tool Count3/5

With only 3 tools, the set feels thin for a server named 'Manus MCP', which suggests a broader AI task management domain. While the tools are well-defined, the count is borderline low, potentially lacking operations like task retrieval, updates, or webhook listing.

Completeness2/5

The tool surface has significant gaps for AI task management: it includes create_task but no get_task, update_task, or delete_task, and for webhooks, it lacks list_webhooks. This incomplete CRUD coverage will likely cause agent failures in common workflows.

Maintenance

ActivityInactive
ResponsivenessSyncing

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/nanameru/Manus-MCP'

If you have feedback or need assistance with the MCP directory API, please join our Discord server