Danube

official

AI Tools Marketplace

View on danubeai.com

What can you do with Danube MCP?

Danube gives AI agents governed access to internal systems and ready-made integrations through one MCP endpoint.

Find the right tool — Search thousands of tools in natural language with search_tools, then inspect schema and readiness via describe_tool.
Run tools and batch calls — Execute any tool by ID or name with execute_tool, or run up to 10 concurrently with batch_execute_tools.
Read stored results — Retrieve a saved result, or a path into it, by its execution_id handle using fetch_result.
Load session context — Read organization and personal context plus available skills at session start with get_context.
Manage credentials and spend — Store API keys with store_credential and set per-call and daily caps via update_spending_limits.
Build and run workflows — Create multi-tool workflows with create_workflow and run them with execute_workflow.

Documentation

Introduction

Governed tool access for AI agents

Danube

What is Danube?

Danube gives AI agents governed access to tools. Your agents call internal systems, databases, and hundreds of ready-made integrations through a single MCP endpoint. They never see a credential, you decide which agent can call what, and every call is logged. Execution can stay inside your own network. Teams can also publish their own tools to the catalog for other agents to use.

[card] Tool Discovery

Search thousands of tools using natural language. Find the right tool for any task.

[card] Unified Access

One MCP connection gives you access to tools email, search, calendar, Slack, GitHub, and more.

[card] Secure Credentials

Manage API keys and OAuth tokens securely. Your credentials are encrypted and never exposed.

[card] Context

Tell every agent what your company does and how you work: organization and personal markdown, read at the start of a session.

[card] Governance

Organizations, roles, tool access policies, and an append-only audit log for every call.

[card] Internal Systems

Describe an internal HTTP service or a read-only database and get governed tools without code.

Get Started

[card] CLI

Install the CLI for agent-first tool execution from the terminal

[card] MCP

Connect via MCP protocol for IDE and desktop integrations

[card] SDK

Use the Python or TypeScript SDK for programmatic access

Available MCP Tools

Danube exposes 33 tools through the MCP server. Connect with ?tools=core (or ?tools=core,workflows) to list only the groups a session needs; see Other clients.

Tool	Description
Discovery
`list_services`	Browse available service providers
`search_tools`	Find tools using semantic search; each result carries `readiness` for your account
`get_service_tools`	Get all tools for a specific service
`describe_tool`	Schema, readiness, 30-day usage with common errors, and your last working parameters for one tool
`get_recommendations`	Get personalized tool recommendations
Execution
`execute_tool`	Run any tool by ID or name with parameters; `fields` keeps only the paths you need
`batch_execute_tools`	Execute up to 10 tools concurrently
`fetch_result`	Read a stored result (or a path into it) back by its `execution_id` handle
`set_parameter_defaults`	Remember parameters your account always passes to a service, filled in when a call omits them
Context
`get_context`	Read your organization's and your personal context, and the skills you can load; call it at the start of a session
`update_context`	Remember something in your personal context (append or replace a document)
Wallet & Spending
`get_wallet_balance`	Check your wallet balance
`get_spending_limits`	View current spending limits
`update_spending_limits`	Configure per-call and daily spending caps
Credentials
`store_credential`	Store an API key or bearer token for a service
Skills
`search_skills`	Search public skills, your organizations' internal skills and your own
`get_skill`	Get full skill content by ID or name, with a `download_url` for the zip
`create_skill`	Create a private skill or an internal one for your organization
`update_skill`	Update an existing skill
`delete_skill`	Delete a skill you own
Workflows
`list_workflows`	List public workflows
`create_workflow`	Create a multi-tool workflow
`update_workflow`	Update an existing workflow
`delete_workflow`	Delete a workflow you own
`execute_workflow`	Execute a multi-tool workflow
`get_workflow_execution`	Get a workflow execution result
Agents
`register_agent`	Register an autonomous agent
`get_agent_info`	Get agent profile and wallet info
`fund_agent_wallet`	Fund an agent's wallet
Ratings & Trust
`submit_rating`	Rate a tool (1-5 stars)
`get_my_rating`	Get your rating for a tool
`get_tool_ratings`	Get aggregated ratings for a tool
`report_tool`	Report a tool that is broken, degraded, returning incorrect output, or timing out

Supported Clients

Danube works with any MCP-compatible client:

[card] Claude Desktop

Anthropic's desktop app

[card] Claude Code CLI

Terminal AI assistant

[card] Codex

OpenAI's desktop app

[card] Codex CLI

OpenAI's terminal agent

[card] Cursor

AI-powered code editor

[card] Windsurf

Codeium's code editor

[card] Hermes Agent

Nous Research's open-source agent

[card] More...

Any MCP client

Get Started

[card] Quickstart

Connect to Danube in under 5 minutes

[card] MCP Clients

Setup guides for Claude Desktop, Cursor, and more

This documentation is built and hosted on Mintlify, a developer documentation platform.