Obot AI

What Obot AI does

Obot AI provides an open-source, Model Context Protocol (MCP) gateway that manages MCP servers, AI Skills, access policies, audit logs, and integrations for AI agents. It offers a unified control plane for enterprises to securely connect, manage, and scale AI agents and tools across their systems, ensuring governance, visibility, and trust. Obot helps organizations move from AI experimentation to real-world adoption by providing a standard foundation for building, running, and scaling MCP servers.

Products

• Obot MCP GatewayA complete solution for securing and managing MCP servers and AI Skills. It acts as a central point of control for how organizations manage and use Model Context Protocol servers, providing a unified gateway for OAuth, access control, security policies, audit logs, discovery, deployment, and AI agent connections.
• Obot SentryA companion agent installed on user workstations for device inventory and local AI activity. It enrolls devices, inventories supported AI clients, MCP servers, skills, and plugins, and submits local tool-call events for auditing.
• LLM GatewayPresents provider-compatible endpoints that external AI clients can use with scoped Obot API keys. It manages upstream provider credentials and applies Model Access Policies before forwarding requests, supporting various LLM providers and recording request status, metadata, token counts, and estimated model cost.

Key capabilities

• OAuth for all MCP servers and Skills
• Access control & identity integration
• Security policies and audit logs
• Discovery and deployment of MCP servers and Skills
• AI agent connections
• Onboard and manage MCP servers and Skills
• Define secure access policies
• Track usage, performance, and compliance
• Help users find and connect to trusted MCPs and Skills
• Central Control Plane for IT
• Universal MCP Hosting & Proxying
• Skills Registry and Discovery
• Instant MCP Connections
• MCP Catalog & Discovery
• MCP Server Management & Integration
• AI Agents & MCP
• MCP Monitoring & Audit Logging
• Admin Setup & Integration with identity providers (GitHub, Google, Okta, Microsoft Entra)
• Catalog Configuration with built-in and custom MCPs
• Access Control & Policy Management for users and teams
• User Discovery & Connection via universal Obot skill
• Intelligent Routing & Hosting for remote and hosted MCPs
• Full Visibility & Control of all interactions
• Host Node.js, Python, and containerized MCP servers on Docker or Kubernetes
• Register remote MCP servers
• Create composite MCP servers
• Manage catalogs through UI or Git-backed sources
• Serve standard MCP Registry API
• Grant access to servers and tools by user or identity-provider group with MCP Access Policies
• Handle MCP OAuth, user and shared credentials, Kubernetes secret bindings, and token exchange
• Inspect, reject, or modify MCP requests and responses with filters
• Restrict domains hosted MCP servers can reach through MCP Server Egress Control
• Index skills from Git repositories and make them available according to Skill Access Policies
• Obot CLI for users and local AI clients to search and install skills
• Centralized Git credentials management
• Model Access Policies for LLM Gateway
• Device Inventory and Local AI Activity with Obot Sentry
• Managed audit hooks for Claude Code, Codex, Cursor, and VS Code
• Audit Logs and Usage for MCP requests, local tool calls, and LLM Gateway requests

Use cases

• Securing and managing MCP servers and AI Skills
• Accelerating AI adoption in enterprises
• Preventing shadow infrastructure, data leaks, and lost productivity in AI integration
• Providing a unified control plane for AI access, policies, auditing, and integrations
• Connecting clients to models and tools
• Managing identities and credentials for AI ecosystems
• Recording activity across centrally hosted services and user workstations
• Defining access control policies for AI resources
• Onboarding new MCPs
• Tracking usage and troubleshooting AI systems
• Hosting and proxying local, remote, and multi-tenant MCPs
• Managing and discovering Skills within an organization
• Connecting MCPs to AI clients via generated URLs or Obot's chat client
• Managing all MCP servers (PostgreSQL-backed, custom internal) through one gateway
• Enabling secure access for AI agents (Claude, Cursor, VSCode) to internal systems
• Monitoring and auditing MCP server and AI agent usage for IT and security teams
• Deploying Obot Gateway Server and connecting to identity providers
• Configuring internal or third-party MCPs via UI or GitOps
• Defining fine-grained access policies for Skills and MCPs
• Enabling user discovery and connection to approved services
• Proxying traffic for remote MCP servers and enforcing policies
• Deploying hosted MCPs in containers on organizational infrastructure
• Governing internal AI ecosystems
• Distributing AI capabilities
• Managing the definition, deployment, discovery, and use of MCP servers
• Granting access to servers and tools by user or identity-provider group
• Handling MCP OAuth and credentials
• Inspecting, rejecting, or modifying MCP requests and responses
• Restricting egress for hosted MCP servers
• Indexing and managing agent skills from Git repositories
• Providing provider-compatible endpoints for external AI clients
• Inventorying AI clients and their configurations on user workstations
• Auditing local tool-call events
• Inspecting operational metadata and accessing stored request/response content for auditors

AI approach

Obot AI provides an open-source MCP (Model Context Protocol) gateway that helps organizations manage, secure, and scale AI agents and tools. It acts as a central control plane for AI access, policies, auditing, and integrations, allowing enterprises to host, proxy, and observe MCP servers. Obot does not build its own foundational AI models but rather provides infrastructure for managing and governing the use of various AI models and tools within an organization.

Tech named: Model Context Protocol (MCP), OAuth, GitOps, PostgreSQL, Claude, Cursor, VSCode, Zapier, AWS, GitHub, Google, Okta, Microsoft Entra, Node.js, Python, Docker, Kubernetes, OpenAI, Anthropic, Amazon Bedrock, Azure OpenAI, Microsoft Foundry, Ollama, LiteLLM, Microsoft Intune

What it says sets it apart

• Open-source platform for managing and governing internal AI ecosystems
• Unified gateway for managing MCP servers, Skills, access policies, audit, and integrations for AI agents
• Central control plane for IT to define access control, onboard MCPs, track usage, and troubleshoot
• Universal MCP hosting and proxying for local, remote, and multi-tenant MCPs
• Comprehensive Skills Registry and Discovery solution
• Built-in MCP Catalog for structure, discovery, and documentation
• Enterprise-grade logging and monitoring for IT and security teams
• Seamless user experience for discovering and connecting to MCPs and Skills
• Supports various identity providers for secure authentication
• GitOps workflows for catalog configuration
• Fine-grained access policies for users and teams
• Intelligent routing and hosting capabilities
• Ability to host Node.js, Python, and containerized MCP servers
• Management of composite MCP servers
• Advanced credential management including OAuth, shared credentials, and Kubernetes secret bindings
• Request and response filtering for MCP traffic
• Egress control for hosted MCP servers
• Centralized management of Git credentials for skill repositories
• LLM Gateway with model access policies and cost tracking
• Device inventory and local AI activity monitoring with Obot Sentry
• Auditing of MCP requests, local tool calls, and LLM Gateway requests with role-based access to content

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.