This is a live Harbor workspace — updated by an AI agent, every change a reviewable diff.Self-host yours →
Live workspace

MCP Ecosystem Tracker

A real Harbor workspace tracking the MCP and AI-tooling ecosystem. Updated on a cadence by a scheduled AI agent. Every change — proposed, reviewed, applied — is logged and visible.

TL;DR

This is a public, read-only Harbor workspace tracking notable builders, MCP servers, and key decisions in the AI-tooling ecosystem. A scheduled agent keeps it current; every AI-made change is logged as an auditable diff at /explore/changes. No account needed to read. You can self-host your own Harbor workspace for free.

This workspace is being set up. Check back soon — it will be populated with real MCP ecosystem data.

State of MCP — 2026

Where the ecosystem stands

The Model Context Protocol (MCP) was introduced by Anthropic in late 2024 as an open standard defining how AI models connect to external tools, data sources, and services. By mid-2026, it has become the primary interoperability layer for AI tooling — the mechanism by which AI assistants discover capabilities, invoke structured operations, and read from real data.

This workspace tracks the builders, servers, clients, and decisions shaping that ecosystem. The table below is a high-level map; the people and project directories carry the sourced detail.

MCP ecosystem — key dimensions
Protocol statusOpen standard (MIT licence). Specification hosted at modelcontextprotocol.io and maintained on GitHub.
Notable clientsClaude Desktop (Anthropic), Cursor, Zed, Cline, Continue, and a growing list of IDE and agent integrations.
Server landscapeHundreds of published implementations catalogued across mcp.so, PulseMCP, Smithery, and the official modelcontextprotocol/servers registry.
Transportstdio for local processes; HTTP + SSE for remote servers. The spec is transport-agnostic by design.
Auth modelBearer token for remote servers. Local servers run as trusted processes within the client.
GovernanceAnthropic maintains the specification; RFCs and issues are open on GitHub at github.com/modelcontextprotocol.

Entries sourced from public ecosystem activity and updated by a scheduled AI agent. Browse the workspace for current, individual profiles.

Frequently asked questions

What is the Model Context Protocol (MCP)?
MCP (Model Context Protocol) is an open standard introduced by Anthropic that defines how AI models connect to external tools, data sources, and services. It lets AI assistants — like Claude — discover and call tools exposed by MCP servers, enabling structured, composable AI workflows.
What is the MCP Ecosystem Tracker?
The MCP Ecosystem Tracker is a public Harbor workspace cataloguing notable builders, MCP servers, clients, and key decisions in the AI-tooling ecosystem. It is updated on a regular cadence by a scheduled AI agent from real public sources.
How is this workspace kept up to date?
A scheduled AI agent reads public sources — release notes, project repositories, and ecosystem announcements — and proposes changes via the Harbor MCP endpoint. Every proposed change is recorded as a reviewable diff and made visible at /explore/changes. No update is hidden.
Can I edit or contribute to this workspace?
This workspace is public and read-only. Anonymous visitors can browse all content. To create your own Harbor workspace — with AI-assisted updates and the same auditable-diff workflow — you can self-host Harbor for free.
What is Harbor?
Harbor is a structured workspace for people, projects, and documents. It combines a team wiki, a knowledge graph, and AI that proposes edits as reviewable diffs. Every change is versioned and auditable. Harbor exposes an MCP endpoint so any MCP-compatible AI client can read and update your workspace.
Audit trail

Every change, visible.

The AI agent proposes each update as a diff. Every patch is approved, recorded, and shown here — an honest, public history of how this workspace stays current.

See all changes →