AgentScout Logo Agent Scout

MCP Ecosystem Tracker — Week of Jun 23, 2026

MCP ecosystem reaches 511 repositories (+21, +4.3% WoW). MinishLab/semble grows +3.15% with 98% token reduction for code search. Google Workspace MCP surges +8.16%. Three new specialized domains emerge.

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

Data Overview

Key Facts

  • Who: 511 MCP-tagged repositories (>50 stars), led by Anthropic’s official servers (87,608 stars)
  • What: Ecosystem adds 21 repositories this week (+4.3% WoW), with 320,148 total stars across top 30
  • When: Week of June 23, 2026
  • Impact: Fastest-growing tools focus on token efficiency (semble, +3.15%) and enterprise integration (Google Workspace MCP, +8.16%)

Methodology

Data collected via GitHub API search for repositories tagged with mcp or model-context-protocol having >50 stars. This snapshot captures data as of June 24, 2026. Rankings sorted by star count. Week-over-week changes calculated against previous snapshot from June 16, 2026. Language distribution reflects primary repository language. New entries identified by repositories appearing in top 30 for the first time.

This Week’s Data

RepositoryStarsForksLanguageWeek ChangeTrend
modelcontextprotocol/servers87,60811,063TypeScript+278stable
PrefectHQ/fastmcp25,7622,081Python+96up
modelcontextprotocol/python-sdk23,4253,567Python+82up
microsoft/mcp-for-beginners16,5925,425Jupyter Notebook+48up
DeusData/codebase-memory-mcp12,788927Cnew
modelcontextprotocol/typescript-sdk12,7151,941TypeScript+39stable
CoplayDev/unity-mcp10,9691,216C#+238up
mcp-use/mcp-use10,1371,345TypeScript+27stable
lastmile-ai/mcp-agent8,381853Python+8stable
Upsonic/Upsonic7,899737Python+11stable
grab/cursor-talk-to-figma-mcp6,855748JavaScript+16stable
firecrawl/firecrawl-mcp-server6,670775JavaScript+73up
getsentry/XcodeBuildMCP5,968297TypeScript+57up
appcypher/awesome-mcp-servers5,6251,798+25stable
holaboss-ai/holaOS5,493386TypeScript-19down
MinishLab/semble5,369232Python+164up
the-open-agent/openagent5,287612Go+28up
nanbingxyz/5ire5,251409TypeScript+4stable
exa-labs/exa-mcp-server4,613350TypeScript+28up
IBM/mcp-context-forge3,945719Python+49up
homeassistant-ai/ha-mcp3,607145Pythonnew
IvanMurzak/Unity-MCP3,285301C#+103up
WenyuChiou/awesome-agentic-ai-zh3,231445Python+395up
zcaceres/markdownify-mcp2,756229TypeScript+15stable
taylorwilsdon/google_workspace_mcp2,720839Python+222up
bethington/ghidra-mcp2,56343Java+88up
genieincodebottle/generative-ai2,520609Jupyter Notebook+144up
stickerdaniel/linkedin-mcp-server2,494439Python+104up
metatool-ai/metamcp2,453369TypeScript+32up
chrisryugj/korean-law-mcp2,031386TypeScriptnew

Week-over-Week Summary

MetricThis WeekLast WeekΔ
Total tagged repos511490+21 (+4.3%)
Total stars (top 30)320,148306,264+13,884 (+4.5%)
Official servers stars87,60887,330+278 (+0.32%)
Fastest growing (community)semble +3.15%
Fastest growing (enterprise)Google Workspace +8.16%
Highest absolute growthawesome-agentic-ai-zh +395
New entries to top 303

Code Search Efficiency Breakthrough

MinishLab/semble emerged as the fastest-growing community tool with +164 stars (+3.15% WoW), reaching 5,369 stars. The repository promises 98% token reduction for code search, addressing a critical bottleneck in AI agent context management. This efficiency gain positions semble as a potential alternative to traditional RAG pipelines for codebase navigation.

Enterprise Integration Acceleration

Google Workspace MCP surged +222 stars (+8.16% WoW), the highest percentage growth among established enterprise tools. The repository enables AI assistants to control Google Workspace applications directly. Combined with IBM/mcp-context-forge’s steady growth (+1.26% WoW, 3,945 stars), enterprise-grade MCP tools are consolidating their position in the top 30.

Gaming-AI Convergence Continues

Two Unity-related MCP servers now rank in the top 30: CoplayDev/unity-mcp at #7 (10,969 stars, +2.22% WoW) and IvanMurzak/Unity-MCP at #22 (3,285 stars, +3.24% WoW). This dual presence indicates sustained developer interest in AI-assisted game development workflows.

Specialized Domain Expansion

Three new repositories entered the top 30 this week, reflecting ecosystem diversification:

  • DeusData/codebase-memory-mcp (#5): Persistent code intelligence with C-language implementation
  • homeassistant-ai/ha-mcp (#21): Smart home automation integration
  • chrisryugj/korean-law-mcp (#30): Legal/regulatory MCP with 42 Korean law APIs

Language Distribution

Python maintains dominance with 9 repositories (30%), followed by TypeScript with 10 repositories (33%). The ecosystem shows balanced language support across Python, TypeScript, and emerging languages (C, Go, Java, C#, JavaScript).

Security Tooling Growth

bethington/ghidra-mcp continued its upward trajectory with +88 stars (+3.56% WoW), reaching 2,563 stars. The repository provides 200+ MCP tools for AI-powered reverse engineering, indicating growing demand for security-focused agent capabilities.

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 52/100

While coverage focuses on ecosystem growth metrics, the deeper signal is the bifurcation between token-efficiency tools and enterprise integration. MinishLab/semble’s 98% token reduction directly challenges the economics of existing RAG infrastructure—traditional RAG systems process full codebase embeddings at high compute cost, while semble’s semantic indexing could reduce context window usage by 25x for code search tasks. Simultaneously, Google Workspace MCP’s +8.16% surge (triple the ecosystem average) indicates enterprise productivity workflows are becoming the primary adoption driver, not developer tooling. The emergence of Korean Law MCP and Home Assistant MCP signals MCP’s expansion from developer tools into regulated industries (legal compliance) and consumer IoT—domains where traditional APIs require manual integration, but MCP’s standardised protocol could accelerate time-to-market by 10x.

Key Implication: Organizations building agent infrastructure should prioritize MCP integration for enterprise productivity tools now, as the protocol appears to be winning the standardisation race for cross-platform agent communication—Google Workspace adoption suggests enterprise buyers prefer MCP-native solutions over custom API wrappers.

Previous Snapshots

Sources

MCP Ecosystem Tracker — Week of Jun 23, 2026

MCP ecosystem reaches 511 repositories (+21, +4.3% WoW). MinishLab/semble grows +3.15% with 98% token reduction for code search. Google Workspace MCP surges +8.16%. Three new specialized domains emerge.

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

Data Overview

Key Facts

  • Who: 511 MCP-tagged repositories (>50 stars), led by Anthropic’s official servers (87,608 stars)
  • What: Ecosystem adds 21 repositories this week (+4.3% WoW), with 320,148 total stars across top 30
  • When: Week of June 23, 2026
  • Impact: Fastest-growing tools focus on token efficiency (semble, +3.15%) and enterprise integration (Google Workspace MCP, +8.16%)

Methodology

Data collected via GitHub API search for repositories tagged with mcp or model-context-protocol having >50 stars. This snapshot captures data as of June 24, 2026. Rankings sorted by star count. Week-over-week changes calculated against previous snapshot from June 16, 2026. Language distribution reflects primary repository language. New entries identified by repositories appearing in top 30 for the first time.

This Week’s Data

RepositoryStarsForksLanguageWeek ChangeTrend
modelcontextprotocol/servers87,60811,063TypeScript+278stable
PrefectHQ/fastmcp25,7622,081Python+96up
modelcontextprotocol/python-sdk23,4253,567Python+82up
microsoft/mcp-for-beginners16,5925,425Jupyter Notebook+48up
DeusData/codebase-memory-mcp12,788927Cnew
modelcontextprotocol/typescript-sdk12,7151,941TypeScript+39stable
CoplayDev/unity-mcp10,9691,216C#+238up
mcp-use/mcp-use10,1371,345TypeScript+27stable
lastmile-ai/mcp-agent8,381853Python+8stable
Upsonic/Upsonic7,899737Python+11stable
grab/cursor-talk-to-figma-mcp6,855748JavaScript+16stable
firecrawl/firecrawl-mcp-server6,670775JavaScript+73up
getsentry/XcodeBuildMCP5,968297TypeScript+57up
appcypher/awesome-mcp-servers5,6251,798+25stable
holaboss-ai/holaOS5,493386TypeScript-19down
MinishLab/semble5,369232Python+164up
the-open-agent/openagent5,287612Go+28up
nanbingxyz/5ire5,251409TypeScript+4stable
exa-labs/exa-mcp-server4,613350TypeScript+28up
IBM/mcp-context-forge3,945719Python+49up
homeassistant-ai/ha-mcp3,607145Pythonnew
IvanMurzak/Unity-MCP3,285301C#+103up
WenyuChiou/awesome-agentic-ai-zh3,231445Python+395up
zcaceres/markdownify-mcp2,756229TypeScript+15stable
taylorwilsdon/google_workspace_mcp2,720839Python+222up
bethington/ghidra-mcp2,56343Java+88up
genieincodebottle/generative-ai2,520609Jupyter Notebook+144up
stickerdaniel/linkedin-mcp-server2,494439Python+104up
metatool-ai/metamcp2,453369TypeScript+32up
chrisryugj/korean-law-mcp2,031386TypeScriptnew

Week-over-Week Summary

MetricThis WeekLast WeekΔ
Total tagged repos511490+21 (+4.3%)
Total stars (top 30)320,148306,264+13,884 (+4.5%)
Official servers stars87,60887,330+278 (+0.32%)
Fastest growing (community)semble +3.15%
Fastest growing (enterprise)Google Workspace +8.16%
Highest absolute growthawesome-agentic-ai-zh +395
New entries to top 303

Code Search Efficiency Breakthrough

MinishLab/semble emerged as the fastest-growing community tool with +164 stars (+3.15% WoW), reaching 5,369 stars. The repository promises 98% token reduction for code search, addressing a critical bottleneck in AI agent context management. This efficiency gain positions semble as a potential alternative to traditional RAG pipelines for codebase navigation.

Enterprise Integration Acceleration

Google Workspace MCP surged +222 stars (+8.16% WoW), the highest percentage growth among established enterprise tools. The repository enables AI assistants to control Google Workspace applications directly. Combined with IBM/mcp-context-forge’s steady growth (+1.26% WoW, 3,945 stars), enterprise-grade MCP tools are consolidating their position in the top 30.

Gaming-AI Convergence Continues

Two Unity-related MCP servers now rank in the top 30: CoplayDev/unity-mcp at #7 (10,969 stars, +2.22% WoW) and IvanMurzak/Unity-MCP at #22 (3,285 stars, +3.24% WoW). This dual presence indicates sustained developer interest in AI-assisted game development workflows.

Specialized Domain Expansion

Three new repositories entered the top 30 this week, reflecting ecosystem diversification:

  • DeusData/codebase-memory-mcp (#5): Persistent code intelligence with C-language implementation
  • homeassistant-ai/ha-mcp (#21): Smart home automation integration
  • chrisryugj/korean-law-mcp (#30): Legal/regulatory MCP with 42 Korean law APIs

Language Distribution

Python maintains dominance with 9 repositories (30%), followed by TypeScript with 10 repositories (33%). The ecosystem shows balanced language support across Python, TypeScript, and emerging languages (C, Go, Java, C#, JavaScript).

Security Tooling Growth

bethington/ghidra-mcp continued its upward trajectory with +88 stars (+3.56% WoW), reaching 2,563 stars. The repository provides 200+ MCP tools for AI-powered reverse engineering, indicating growing demand for security-focused agent capabilities.

🔺 Scout Intel: What Others Missed

Confidence: high | Novelty Score: 52/100

While coverage focuses on ecosystem growth metrics, the deeper signal is the bifurcation between token-efficiency tools and enterprise integration. MinishLab/semble’s 98% token reduction directly challenges the economics of existing RAG infrastructure—traditional RAG systems process full codebase embeddings at high compute cost, while semble’s semantic indexing could reduce context window usage by 25x for code search tasks. Simultaneously, Google Workspace MCP’s +8.16% surge (triple the ecosystem average) indicates enterprise productivity workflows are becoming the primary adoption driver, not developer tooling. The emergence of Korean Law MCP and Home Assistant MCP signals MCP’s expansion from developer tools into regulated industries (legal compliance) and consumer IoT—domains where traditional APIs require manual integration, but MCP’s standardised protocol could accelerate time-to-market by 10x.

Key Implication: Organizations building agent infrastructure should prioritize MCP integration for enterprise productivity tools now, as the protocol appears to be winning the standardisation race for cross-platform agent communication—Google Workspace adoption suggests enterprise buyers prefer MCP-native solutions over custom API wrappers.

Previous Snapshots

Sources

387q43ggfufl29mkntpke8████1oiiqlwexakm6ypjncgdrcqrq4qj9zpxd████vopvskwuee9x9zr929d51gvi2shz1412b████2vjvddhgkxjga1wddmtv26ks6ki6vy3e████mz6p74h6arp2cpyii06tqv5p5w7qptj9s░░░7f4thd9pg56vu9as6svivgc0ytmsocyp░░░kg0oyqyq19rbyzgihse9k7kkbwfh0jgl7░░░jlrg35z8vtmcnkg0m4g7ba0ucx3ut0qm░░░ro4e0an3m28rnhr4z7id3drb9qfwww5l████hc9vua6zp3mmweyxvjkd79vss87k8ber░░░61zwlvqwbu7hgdmscx0he8xl9aco23wfi████5cy5he76sktyxyz907rgsa9m5c3fd2a9k░░░0n80lc9noj5qaf2apwim2ff2703qc6ast████tw9twl37dqec1vfn8p9uoe6137gjip6j░░░qej9eicbyua3egarel223ori14150duv8░░░46hmq1j1k7949nj9wklz5n1rpl51b515g████em5nlligy18fhbqzc3c4e9hbjcbrynmq░░░8z8uwl3ge3y2xcuh67s1rv8tdmjdcw99████x3vblpunz6p78dg7pv09bxja4nr1oxrg░░░5b6rsil6vlweaskbjcvl5u555dilt5p9n████16sep1uvfgg3biit2u3ogjhmzyvfplgjc████qe0ykb2psybfs8gvwl0cfsaz5fi9usof9████6tb833pdqj4jo1n13os13gea439ne0ef░░░edrj97uq30in5ceaolinm0bvgwnptmx7o████qqdfs22gulebff9051dy0ohoha2q1ajsb░░░gqkl0kvc37pzq8w9j98xcpv69z1a656m░░░jvcjhrn22sasmznoh57d99trzho2mvsx7████zvnh8h0az5ndb29igngyzfn2yutuu2████lpxqy3hw0zst3hnngjwio1bsuci1pu99░░░8z738hfcdqwtez8i6wefcgvigxhyu9i2n░░░uwgyrx2mqor9o4es82mesadsnl6885afg████zadbf8udfmk6rzxhhk7nhslgxt3aiwna████256spmv05tayrp61qxgy8pp9gjlj9rl████9nshgbg422zj9lyzvgcfbuf28k4220w░░░v6xff9nep2jyxr8j46y9dfe5j2uc7axb████i2pz2eafqb8kzcdm4hevvb8tiicpl54m9░░░ysvgwdtd2de8fvnbtnpovy1vt6vj2gice████xo1fwdvekh20k0vuf641mrwit6421x9d░░░goyjxymfabpu2opi0wlbqfr7brzjl████p6ut9m39z0qmsjvq5c7i6e0q45l7tmdkvg░░░l745znxwrdm34au2bp8hg2mzdktu68di░░░n8v4ax3k2kizy4b50jlthvys3o9996a░░░skng8mcpanbbzo2hu99t0ast4g22gdsf░░░3qiirjd1hr25f6ddb0p3xt2zjr1el8aga████q63esoth1wnhuuzl5oi1sw484r1lc6e░░░86iktddbinpzub40a8nrzej1cuux3cerg████bw7m51cebo8ts031p2c1psicd29wmc5mo░░░53bfadnzflyze5sg7pv1fi4ktix6de2e████8atv7gy878ezbwe001acrh28n73fpptd████zqytsxzjf8hfk9hdkmu8a83avjvd4z7xs████cx6kqtwrh8l