MCP & AI Agent Integration

Connect AI assistants directly to your Nestr workspace through the Model Context Protocol (MCP). AI agents can read your governance structure, create and process tensions, manage work, and act from roles, making powerful agentic workflows possible within your self-organised team.

What is the Model Context Protocol (MCP)?

MCP is an open standard that lets AI assistants connect directly to external tools and data sources. When connected to Nestr, an AI agent can read your full organisational structure, understand who holds which role, and take actions within the boundaries of those roles.

How does MCP integration work in Nestr?

Once connected, your AI assistant has access to Nestr's full functionality through natural language. It can query roles and circles, create tensions, manage projects, process governance proposals, and update work items. The agent acts within the governance boundaries you have defined, keeping its behaviour accountable and traceable in the audit trail.

What can AI agents do in Nestr via MCP?

• Read the full organisational structure: circles, roles, accountabilities, and domains
• Create and resolve tensions on behalf of any role they fill
• Create, update, and complete projects and todos
• Participate in governance proposals
• Post to circle feeds and send notifications
• Access insights and metrics data
• Almost everything you can do through the UI

Connectors: the same standard, pointing outwards

MCP brings outside assistants in. Connectors do the reverse, giving Nestr's own AI colleagues governed access out to the systems your organisation already runs on: Gmail and Outlook, Google Drive and OneDrive, Slack and Teams, GitHub, GitLab, Jira and Confluence, Linear, Notion, Asana, Monday.com, ClickUp, Sentry, Stripe, Xero, QuickBooks, Moneybird, HubSpot, Intercom, Dropbox, Box, Zapier and Make. You can register a custom connector against any API or MCP server you run yourself, including internal systems.

Access is a governance decision rather than a shared password. Admins decide which data sources exist and how broadly they may be owned; whoever may edit a role's domain decides which roles get them. Credentials are held in an external vault, never travel to the agent, and every use is recorded against the role that made it.