AgentScout Logo Agent Scout

MCP Ecosystem Weekly Tracker

Weekly snapshot of the Model Context Protocol ecosystem: 345 repositories tracked, FastMCP leads community growth with 24,535 stars, Python SDK outpaces TypeScript by 85%.

AgentScout Β· Β· Β· 5 min read
#mcp #model-context-protocol #ecosystem #weekly-tracker #anthropic #ai-agents
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

Data Overview

  • Last Updated: 2026-04-15
  • Update Frequency: Weekly
  • Primary Sources: GitHub API (MCP topic repositories), Awesome MCP Servers List

The Model Context Protocol (MCP) ecosystem continues its expansion with 345 repositories exceeding 50 stars out of 5,813 total tagged repositories. This weekly tracker monitors the top projects, language distribution, and emerging trends across the MCP landscape.

Methodology

Data Collection:

  • GitHub API queries for repositories tagged with mcp and model-context-protocol topics
  • Filtered to repositories with >50 stars for quality threshold
  • Data snapshot taken on 2026-04-15

Validation Criteria:

  • Repository must have mcp or model-context-protocol as a topic tag
  • Minimum 50 stars to qualify for tracking
  • Active development assessed via Last Updated field

Column Definitions:

ColumnDefinition
RepositoryGitHub owner/repo name
StarsTotal GitHub stars at collection time
ForksTotal GitHub forks at collection time
LanguagePrimary programming language
TrendClassification of project type/trajectory

Current Data: Top 30 MCP Repositories

RepositoryStarsForksLanguageTrend
modelcontextprotocol/servers83,74110,382TypeScriptofficial-core
PrefectHQ/fastmcp24,5351,928Pythonrising-star
modelcontextprotocol/python-sdk22,6343,321Pythonofficial-sdk
microsoft/mcp-for-beginners15,8725,171Jupyter Notebookeducational
modelcontextprotocol/typescript-sdk12,1811,775TypeScriptofficial-sdk
mcp-use/mcp-use9,7731,242TypeScriptrising
CoplayDev/unity-mcp8,482968C#game-dev
lastmile-ai/mcp-agent8,265830Pythonagent-framework
Upsonic/Upsonic7,825726Pythonenterprise
grab/cursor-talk-to-figma-mcp6,651721JavaScriptdesign-tool
firecrawl/firecrawl-mcp-server6,050680JavaScriptweb-tool
appcypher/awesome-mcp-servers5,3931,219-directory
getsentry/XcodeBuildMCP5,196245TypeScriptios-dev
nanbingxyz/5ire5,155404TypeScriptdesktop-client
casibase/casibase4,493526Goenterprise-platform
exa-labs/exa-mcp-server4,242318TypeScriptsearch-tool
IBM/mcp-context-forge3,569624Pythonenterprise-gateway
zcaceres/markdownify-mcp2,584213TypeScriptutility
holaboss-ai/holaOS2,486271TypeScriptagent-os
genieincodebottle/generative-ai2,215542Jupyter Notebookeducational
metatool-ai/metamcp2,212324TypeScriptmeta-tool
jamubc/gemini-mcp-tool2,155182TypeScriptllm-bridge
IvanMurzak/Unity-MCP2,136195C#game-dev
taylorwilsdon/google_workspace_mcp2,112632Pythonworkspace
zinja-coder/jadx-ai-mcp1,765182Javareverse-engineering
ravitemer/mcphub.nvim1,76083Luaeditor-plugin
stacklok/toolhive1,716207Goenterprise
doobidoo/mcp-memory-service1,668249Pythonmemory
CoderGamester/mcp-unity1,588201C#game-dev
stickerdaniel/linkedin-mcp-server1,547284Pythonsocial

Ecosystem Metrics

Language Distribution (Top 30)

LanguageCountPercentage
Python930%
TypeScript930%
C#310%
JavaScript26.7%
Go26.7%
Jupyter Notebook26.7%
Other516.7%

Aggregate Statistics

MetricValue
Total tagged repositories5,813
Repositories with >50 stars345
Total stars (Top 30)249,742
New repositories this weekN/A (first detailed snapshot)

1. Python SDK Dominates TypeScript SDK

The official Python SDK (22,634 stars) now leads the TypeScript SDK (12,181 stars) by an 85% margin. This signals stronger developer preference for Python in the MCP ecosystem, likely driven by the AI/ML community’s existing Python tooling.

2. FastMCP Emerges as Leading Community Framework

FastMCP (PrefectHQ) has reached 24,535 stars, positioning itself as the most popular community-built MCP server framework. Its tagline β€œThe fast, Pythonic way to build MCP servers” resonates with developers seeking simplicity over official SDK complexity.

3. Unity Game Development Shows Strong Adoption

Three dedicated Unity MCP projects appear in the top 30:

  • CoplayDev/unity-mcp: 8,482 stars
  • IvanMurzak/Unity-MCP: 2,136 stars
  • CoderGamester/mcp-unity: 1,588 stars

This represents 10% of top projects focusing on game development integration.

4. Enterprise MCP Platforms Gaining Traction

Enterprise-focused solutions are emerging:

  • IBM/mcp-context-forge: AI Gateway, registry, and proxy for MCP/A2A/REST APIs
  • Stacklok/toolhive: Enterprise-grade platform for running MCP servers
  • casibase/casibase: Enterprise-level AI knowledge base and MCP management

5. Web Scraping and Search Tools Show Strong Demand

Firecrawl (6,050 stars) and Exa (4,242 stars) MCP servers indicate significant interest in web data extraction capabilities for AI agents.

6. Desktop MCP Clients Emerging

5ire (5,155 stars) demonstrates demand for cross-platform desktop AI assistants that serve as MCP clients, extending beyond terminal-based tooling.

7. Memory and Knowledge Graph Services Essential

Memory/persistence MCP services like mcp-memory-service (1,668 stars) address the critical need for agent context retention across sessions.

Awesome MCP Servers Categories

The community-curated list (punkpeye/awesome-mcp-servers) organizes projects into 19 categories:

CategoryFocus Area
AggregatorsMulti-server management and routing
Art & CultureMuseum, gallery, cultural APIs
Aerospace & AstrodynamicsSpace-related data services
Browser AutomationWeb scraping, browser control
Biology Medicine & BioinformaticsHealthcare and life sciences
Cloud PlatformsAWS, GCP, Azure integrations
Code ExecutionSandboxed code running
Coding AgentsAI-assisted development
Command LineCLI tool integrations
CommunicationSlack, Discord, email
DatabasesSQL, NoSQL connections
Data PlatformsAnalytics, warehousing
Developer ToolsIDE extensions, utilities
File SystemsLocal and cloud storage
Finance & FintechMarket data, banking APIs
GamingGame engine integrations
Knowledge & MemoryVector stores, knowledge graphs
Search & Data ExtractionWeb search, crawling
SecurityAuth, encryption, scanning
Social MediaTwitter, LinkedIn, etc.

Notable Aggregators

ProjectDescription
APIFoldTurn any REST API into a hosted MCP server
metatool-ai/metamcpMCP Aggregator in one Docker container
mcp-use/mcp-useFullstack MCP framework for ChatGPT/Claude apps
PipedreamHQ/pipedreamConnect with 2,500+ APIs via MCP

πŸ”Ί Scout Intel: What Others Missed

Confidence: high | Novelty Score: 65/100

While most MCP coverage focuses on Anthropic’s official servers and SDKs, the ecosystem data reveals a deeper structural shift: community frameworks are outpacing official tooling in developer adoption. FastMCP’s 24,535 stars versus the official Python SDK’s 22,634 stars suggests developers prefer opinionated, simplified interfaces over flexible official SDKs. The 85% gap between Python and TypeScript SDKs indicates Python has become the de facto language for MCP development, aligning with broader AI/ML tooling preferences.

Key Implication: Enterprise teams evaluating MCP should prioritize Python-first implementations and consider FastMCP for rapid prototyping, while monitoring IBM Context Forge and Stacklok ToolHive for production-grade deployment patterns.

Changelog

DateChangeDetails
2026-04-15addedInitial weekly snapshot with top 30 repositories and ecosystem metrics

Sources

MCP Ecosystem Weekly Tracker

Weekly snapshot of the Model Context Protocol ecosystem: 345 repositories tracked, FastMCP leads community growth with 24,535 stars, Python SDK outpaces TypeScript by 85%.

AgentScout Β· Β· Β· 5 min read
#mcp #model-context-protocol #ecosystem #weekly-tracker #anthropic #ai-agents
Analyzing Data Nodes...
SIG_CONF:CALCULATING
Verified Sources

Data Overview

  • Last Updated: 2026-04-15
  • Update Frequency: Weekly
  • Primary Sources: GitHub API (MCP topic repositories), Awesome MCP Servers List

The Model Context Protocol (MCP) ecosystem continues its expansion with 345 repositories exceeding 50 stars out of 5,813 total tagged repositories. This weekly tracker monitors the top projects, language distribution, and emerging trends across the MCP landscape.

Methodology

Data Collection:

  • GitHub API queries for repositories tagged with mcp and model-context-protocol topics
  • Filtered to repositories with >50 stars for quality threshold
  • Data snapshot taken on 2026-04-15

Validation Criteria:

  • Repository must have mcp or model-context-protocol as a topic tag
  • Minimum 50 stars to qualify for tracking
  • Active development assessed via Last Updated field

Column Definitions:

ColumnDefinition
RepositoryGitHub owner/repo name
StarsTotal GitHub stars at collection time
ForksTotal GitHub forks at collection time
LanguagePrimary programming language
TrendClassification of project type/trajectory

Current Data: Top 30 MCP Repositories

RepositoryStarsForksLanguageTrend
modelcontextprotocol/servers83,74110,382TypeScriptofficial-core
PrefectHQ/fastmcp24,5351,928Pythonrising-star
modelcontextprotocol/python-sdk22,6343,321Pythonofficial-sdk
microsoft/mcp-for-beginners15,8725,171Jupyter Notebookeducational
modelcontextprotocol/typescript-sdk12,1811,775TypeScriptofficial-sdk
mcp-use/mcp-use9,7731,242TypeScriptrising
CoplayDev/unity-mcp8,482968C#game-dev
lastmile-ai/mcp-agent8,265830Pythonagent-framework
Upsonic/Upsonic7,825726Pythonenterprise
grab/cursor-talk-to-figma-mcp6,651721JavaScriptdesign-tool
firecrawl/firecrawl-mcp-server6,050680JavaScriptweb-tool
appcypher/awesome-mcp-servers5,3931,219-directory
getsentry/XcodeBuildMCP5,196245TypeScriptios-dev
nanbingxyz/5ire5,155404TypeScriptdesktop-client
casibase/casibase4,493526Goenterprise-platform
exa-labs/exa-mcp-server4,242318TypeScriptsearch-tool
IBM/mcp-context-forge3,569624Pythonenterprise-gateway
zcaceres/markdownify-mcp2,584213TypeScriptutility
holaboss-ai/holaOS2,486271TypeScriptagent-os
genieincodebottle/generative-ai2,215542Jupyter Notebookeducational
metatool-ai/metamcp2,212324TypeScriptmeta-tool
jamubc/gemini-mcp-tool2,155182TypeScriptllm-bridge
IvanMurzak/Unity-MCP2,136195C#game-dev
taylorwilsdon/google_workspace_mcp2,112632Pythonworkspace
zinja-coder/jadx-ai-mcp1,765182Javareverse-engineering
ravitemer/mcphub.nvim1,76083Luaeditor-plugin
stacklok/toolhive1,716207Goenterprise
doobidoo/mcp-memory-service1,668249Pythonmemory
CoderGamester/mcp-unity1,588201C#game-dev
stickerdaniel/linkedin-mcp-server1,547284Pythonsocial

Ecosystem Metrics

Language Distribution (Top 30)

LanguageCountPercentage
Python930%
TypeScript930%
C#310%
JavaScript26.7%
Go26.7%
Jupyter Notebook26.7%
Other516.7%

Aggregate Statistics

MetricValue
Total tagged repositories5,813
Repositories with >50 stars345
Total stars (Top 30)249,742
New repositories this weekN/A (first detailed snapshot)

1. Python SDK Dominates TypeScript SDK

The official Python SDK (22,634 stars) now leads the TypeScript SDK (12,181 stars) by an 85% margin. This signals stronger developer preference for Python in the MCP ecosystem, likely driven by the AI/ML community’s existing Python tooling.

2. FastMCP Emerges as Leading Community Framework

FastMCP (PrefectHQ) has reached 24,535 stars, positioning itself as the most popular community-built MCP server framework. Its tagline β€œThe fast, Pythonic way to build MCP servers” resonates with developers seeking simplicity over official SDK complexity.

3. Unity Game Development Shows Strong Adoption

Three dedicated Unity MCP projects appear in the top 30:

  • CoplayDev/unity-mcp: 8,482 stars
  • IvanMurzak/Unity-MCP: 2,136 stars
  • CoderGamester/mcp-unity: 1,588 stars

This represents 10% of top projects focusing on game development integration.

4. Enterprise MCP Platforms Gaining Traction

Enterprise-focused solutions are emerging:

  • IBM/mcp-context-forge: AI Gateway, registry, and proxy for MCP/A2A/REST APIs
  • Stacklok/toolhive: Enterprise-grade platform for running MCP servers
  • casibase/casibase: Enterprise-level AI knowledge base and MCP management

5. Web Scraping and Search Tools Show Strong Demand

Firecrawl (6,050 stars) and Exa (4,242 stars) MCP servers indicate significant interest in web data extraction capabilities for AI agents.

6. Desktop MCP Clients Emerging

5ire (5,155 stars) demonstrates demand for cross-platform desktop AI assistants that serve as MCP clients, extending beyond terminal-based tooling.

7. Memory and Knowledge Graph Services Essential

Memory/persistence MCP services like mcp-memory-service (1,668 stars) address the critical need for agent context retention across sessions.

Awesome MCP Servers Categories

The community-curated list (punkpeye/awesome-mcp-servers) organizes projects into 19 categories:

CategoryFocus Area
AggregatorsMulti-server management and routing
Art & CultureMuseum, gallery, cultural APIs
Aerospace & AstrodynamicsSpace-related data services
Browser AutomationWeb scraping, browser control
Biology Medicine & BioinformaticsHealthcare and life sciences
Cloud PlatformsAWS, GCP, Azure integrations
Code ExecutionSandboxed code running
Coding AgentsAI-assisted development
Command LineCLI tool integrations
CommunicationSlack, Discord, email
DatabasesSQL, NoSQL connections
Data PlatformsAnalytics, warehousing
Developer ToolsIDE extensions, utilities
File SystemsLocal and cloud storage
Finance & FintechMarket data, banking APIs
GamingGame engine integrations
Knowledge & MemoryVector stores, knowledge graphs
Search & Data ExtractionWeb search, crawling
SecurityAuth, encryption, scanning
Social MediaTwitter, LinkedIn, etc.

Notable Aggregators

ProjectDescription
APIFoldTurn any REST API into a hosted MCP server
metatool-ai/metamcpMCP Aggregator in one Docker container
mcp-use/mcp-useFullstack MCP framework for ChatGPT/Claude apps
PipedreamHQ/pipedreamConnect with 2,500+ APIs via MCP

πŸ”Ί Scout Intel: What Others Missed

Confidence: high | Novelty Score: 65/100

While most MCP coverage focuses on Anthropic’s official servers and SDKs, the ecosystem data reveals a deeper structural shift: community frameworks are outpacing official tooling in developer adoption. FastMCP’s 24,535 stars versus the official Python SDK’s 22,634 stars suggests developers prefer opinionated, simplified interfaces over flexible official SDKs. The 85% gap between Python and TypeScript SDKs indicates Python has become the de facto language for MCP development, aligning with broader AI/ML tooling preferences.

Key Implication: Enterprise teams evaluating MCP should prioritize Python-first implementations and consider FastMCP for rapid prototyping, while monitoring IBM Context Forge and Stacklok ToolHive for production-grade deployment patterns.

Changelog

DateChangeDetails
2026-04-15addedInitial weekly snapshot with top 30 repositories and ecosystem metrics

Sources

v8wfq5r139gmthsmyla5wfβ–ˆβ–ˆβ–ˆβ–ˆsp1urc85qqc1iby4cmq01injf0j49lbβ–‘β–‘β–‘xhauzakujdjacb3pqfijaax28b95av4β–ˆβ–ˆβ–ˆβ–ˆveqoc76l61aqlcbxw6efmoll8lb3uj3qaβ–ˆβ–ˆβ–ˆβ–ˆ6oxnntffi4xvglj705zcrriq1zp0lwsrβ–ˆβ–ˆβ–ˆβ–ˆ5vx3egnhd2n0fy0lqchgcscx5k8j1wpcβ–‘β–‘β–‘t0lpjc6sm6qiyczankfs6nd3raw7776β–‘β–‘β–‘c466kq5f45pjwp2jtj9smk8d8javf1qvhβ–ˆβ–ˆβ–ˆβ–ˆb9fkldn720ovfqoxbs0vwn5iuthmqcdβ–ˆβ–ˆβ–ˆβ–ˆ9jrhpxcg5h85uilunu1hu7adr6r8dcxbaβ–‘β–‘β–‘nlqjdzmtkgcc6r033fmblmrc2wra1908β–‘β–‘β–‘t0u8oziwzw9f3sc8dtj2i7wd9trjhmaeβ–‘β–‘β–‘s15pd3ttbxr3vrh8yia75376xtoq1kufnβ–‘β–‘β–‘16ly5srt06ou09cjso65nh2b9q2u60l3qβ–‘β–‘β–‘72otxnm9ttxqb0elat5ciepb6icepy8rsβ–‘β–‘β–‘v6dli7dnhsm1dfkt4otdx5vnk6kijq4zsβ–ˆβ–ˆβ–ˆβ–ˆuvwku2hmjnoswxbgdfwczqedmj3wvfqβ–‘β–‘β–‘ty64snv0oxjyrwv0ik8r8dpfmrxa8il7β–‘β–‘β–‘y2ycjvae47sgjpg39o6ty6vk8xd7dakibβ–‘β–‘β–‘82uwxyht9pr9ij9yc4tr4vz4inhfulqgβ–ˆβ–ˆβ–ˆβ–ˆj9zoqwt1dgj1geg2tsznuk65m1va9n2thβ–‘β–‘β–‘zgpzdn3sckfgntri2p1ioqbk2yro1wldβ–‘β–‘β–‘o0w3es8kx7wslt7ddc19l3iqail7de8oβ–‘β–‘β–‘akeo1xeky45qkohc1z66sdalfb3ow1pqβ–ˆβ–ˆβ–ˆβ–ˆ5weulf84plj9toees50zcqrr6ddy7r3aβ–ˆβ–ˆβ–ˆβ–ˆd6ry0flhkrleo8paosd5cck9ps3tggmfβ–‘β–‘β–‘yvot901oi62k8rcomh7l3za1mld8g4wβ–‘β–‘β–‘usbtly4hahrs6n07xuykzpfy5zyfen1hβ–‘β–‘β–‘ueqwadhpassbmryvo0t177js20lx2tyiβ–ˆβ–ˆβ–ˆβ–ˆiikurxmgfprc5e3n0ayivbk6o0ns3cweβ–‘β–‘β–‘4k0uwqobs83u5t5a076r9syue79nue5β–‘β–‘β–‘pnqvqfe0s7gv2xqmyf8ha3clxsn29facβ–‘β–‘β–‘r0cxxa6snbcfc7fcxj774tfkdnkq62fnβ–ˆβ–ˆβ–ˆβ–ˆ4c6ap9snr5wphn0fehi36smcw8s3ma99oβ–‘β–‘β–‘9a507drlr1ftevnisec35lehaub9biyqβ–ˆβ–ˆβ–ˆβ–ˆcervac1k1oszmvrp0u8i8ypud7giaz8sβ–‘β–‘β–‘538fk63ipe68s4h5i6yqukyiwwsg7cmnbβ–ˆβ–ˆβ–ˆβ–ˆsu83chnf2r9oavj3x97vqzxy3w469dfβ–‘β–‘β–‘vd3yvylxd0m72hnkksnhc9cdlkhik99bnβ–ˆβ–ˆβ–ˆβ–ˆeyntrgypgvge89ir66czre7ap4fupbk6tβ–‘β–‘β–‘lpivho4w5bksevh0vymwtomv2aa9hglwnβ–‘β–‘β–‘ak8xy9qiz2cxtwcmsiyjqofow357g0oβ–ˆβ–ˆβ–ˆβ–ˆ4r6y47vd07jl5xw846bivqw2pnc9zlmbqβ–ˆβ–ˆβ–ˆβ–ˆc4aslr8pkg8achth0wenvtf3kn1q3d0skβ–‘β–‘β–‘bnw1mytdqnp9mcx70ztqivsxkh9f8kbβ–‘β–‘β–‘qdiidsj2rteduijmfj2rtha7c3uy002ebβ–‘β–‘β–‘tvgsyvcyvciullih7h8v2erb3i6m37rffβ–‘β–‘β–‘opna3n0c6tgm4dgbsbcpfj0s9zz724botβ–‘β–‘β–‘g1wv1t6tdqihv7b68s6s1ef6hvv8ewaβ–ˆβ–ˆβ–ˆβ–ˆpmjql4jiygvw966aft2bf3unvr0g4gnaβ–‘β–‘β–‘qevww7d6oms