AgentSidekick

Most crowded categories in the agent ecosystem

Saturation ranking of 454 taxonomy segments and 111 integrations: size, launch momentum and concentration of attention.

Executive summary

  • Agent Commerce is the largest top-level category with 3,621 products; AI Coding Agents follows with 3,366.

  • 69 of 454 taxonomy segments launch faster than the catalogue as a whole (labelled outpacing or emerging); 176 lag it or are slowing.

  • In 139 of 441 segments with at least five repositories, a single project holds half or more of all GitHub stars.

  • The most crowded specific capability is SEO Research with 1,073 products, 735 of them launched in the last 90 days.

  • Among integrations, YouTube is mentioned by the most products (574).

Every figure above is computed from the database DER

Methodology

How this report was produced, and its limits

Generated on 2026-11-21 from the AgentSidekick database. The latest completed ingestion step finished at 2026-11-21 13:23 UTC. Tables queried: market_segments, taxonomy_nodes, integrations. The catalogue holds 73,820 active products, of which 9,469 meet the data-completeness bar and have a profile page.

Segments were last calculated at 2026-11-21 13:23 UTC.

Classification method: product type comes from the source (registry entries are MCP servers by definition) or is inferred from repository topics; categories, capabilities, integrations and industries come from keyword rules in the editable taxonomy, so they are derived, not checked by a person. Curated records were researched against the official sources listed on each record.

Known limits:

  • Segment size counts every active product matched to the segment, profiled or not. A product can sit in several segments.
  • Top share = the share of the segment's GitHub stars held by its most-starred project; HHI = the sum of squared star shares (0 to 1; above 0.25 is conventionally "highly concentrated"). Both only use products with a repository, and stars are attention, not market share.
  • Labels compare a segment's size with peers at the same level (80th percentile and up = crowded, 30th and below = thin) and its launch momentum with the catalogue's: momentum = 50 + 50 × tanh(ln(segment ratio ÷ catalogue ratio)), where ratio = (launches in the last 90 days + 1) ÷ (launches in the 90 before + 1). 62 or more = outpacing the catalogue, 38 or less = lagging or slowing. Fewer than five launches across both windows = no momentum verdict. Because the whole catalogue is growing fast, a segment can add products every month and still be labelled lagging.
  • Launch dates are the first publish date in the source registry (official MCP Registry) or the GitHub repository creation date. They are not product announcements, and a re-published entry can look like a new launch.
  • GitHub stars measure attention on a repository, not installs, active users or revenue. Several products can share one repository.
  • Only 1 daily metrics snapshot(s) exist in product_metrics, so growth is measured from launch counts, not from changes in stars or downloads.
  • Categories, capabilities and integrations are assigned by keyword rules over names and descriptions (derived). Products that do not name a capability are missed, and generic words can produce false matches.

Note. Claim labels: source = stated by the publisher or registry; curated = checked by us against cited sources; derived = computed by a documented rule; heuristic = a rule of thumb with no measurement behind it. Full method: methodology.

How segments are labelled

Every taxonomy node and integration is scored on two axes: size relative to its peers, and launch momentum. "Crowded" means the segment is in the top fifth by size at its level; the suffix says whether its launches are growing faster or slower than launches across the whole catalogue. 17 segments are crowded and still outpacing the catalogue, and 120 are thin. A crowded label is a description of supply, not a statement that there is no room: a large segment with a dispersed star distribution can still reward a better product.

Taxonomy segments by labelDER

Unit: segments. Top 9 shown.

Taxonomy segments by labelUnit: segments. Top 9 shown.. Highest: Lagging the catalogue, 125 segments.Lagging the catalogue125Thin120Established62Crowded, slowing51Outpacing the catalogue35Crowded26Emerging17Crowded, outpacing the catalogue17Empty1
Show the data as a table
Taxonomy segments by label. Unit: segments. Top 9 shown.
segments
Lagging the catalogue125
Thin120
Established62
Crowded, slowing51
Outpacing the catalogue35
Crowded26
Emerging17
Crowded, outpacing the catalogue17
Empty1

Top-level categories (level 2)

Agent Commerce is the largest level-2 segment with 3,621 products (2,284 vs 782 launches in the last two 90-day windows). 1 of the 20 largest are crowded and launching faster than the catalogue as a whole. The most concentrated among them is Finance & Accounting, where the leading project holds 33% of GitHub stars (HHI 0.16). The least concentrated is Agent Frameworks & Orchestration (top share 5%, HHI 0.02), where no single project dominates.

Largest level-2 categoriesDER

Largest level-2 categories
SegmentProductsWith repoNew 90dPrev 90dTop shareHHILabel
Agent Commerce3,6211,7092,28478217%0.08Crowded
AI Coding Agents3,3662,8241,3346957%0.02Crowded, slowing
Communication & Messaging2,6871,3871,74346833%0.14Crowded
Public & Real-world Data2,0111,3711,11362223%0.11Crowded, slowing
Social & Streaming Platforms2,0011,0131,39133922%0.08Crowded, outpacing the catalogue
Marketing1,9441,1611,28935823%0.10Crowded
Cloud Providers1,7771,27790334213%0.04Crowded
Retrieval & RAG1,7561,4855984205%0.02Crowded, slowing
Crypto & Blockchain1,61497485336721%0.09Crowded, slowing
Files & Documents1,61396395833320%0.07Established
Finance & Accounting1,6005781,19727033%0.16Outpacing the catalogue
Agent Security1,4811,1106503689%0.03Lagging the catalogue
Web Data1,4701,06496725820%0.08Established
Agent Runtime & Deployment1,4611,1706143949%0.04Lagging the catalogue
DevOps & SRE1,3921,1475183134%0.02Lagging the catalogue
Research & Knowledge1,27992159829111%0.04Lagging the catalogue
E-commerce1,24265577726820%0.10Established
Generative Media1,23293170528610%0.04Lagging the catalogue
Markets & Trading1,20973768429728%0.14Lagging the catalogue
Agent Frameworks & Orchestration1,1921,0403112045%0.02Lagging the catalogue

Subcategories (level 3)

Agent Payments & Monetization is the largest level-3 segment with 3,621 products (2,284 vs 782 launches in the last two 90-day windows). 6 of the 20 largest are crowded and launching faster than the catalogue as a whole. The most concentrated among them is Sales Automation, where the leading project holds 50% of GitHub stars (HHI 0.28). The least concentrated is Retrieval-Augmented Generation (top share 6%, HHI 0.02), where no single project dominates.

Largest level-3 subcategoriesDER

Largest level-3 subcategories
SegmentProductsWith repoNew 90dPrev 90dTop shareHHILabel
Agent Payments & Monetization3,6211,7092,28478217%0.08Crowded
Email1,7188341,16930030%0.14Crowded
Marketing Automation1,6709711,13528429%0.12Crowded, outpacing the catalogue
Finance Automation1,6005781,19727033%0.16Crowded, outpacing the catalogue
Context Engineering1,4661,2924803509%0.03Crowded, slowing
Social Networks1,4637461,01325115%0.06Crowded, outpacing the catalogue
Document Processing1,40581588727929%0.12Crowded
Blockchain Data1,35282871631722%0.09Crowded, slowing
Edge & Serverless Platforms1,28093472423819%0.07Crowded
Retrieval-Augmented Generation1,1931,0273942776%0.02Crowded, slowing
Public Datasets1,12684966834718%0.10Crowded, slowing
Code Review1,09494654622535%0.24Crowded, slowing
Chat & Messaging1,04158062418138%0.18Crowded
Web Scraping92274463314724%0.11Crowded, outpacing the catalogue
Video & Audio Platforms87946661615732%0.13Crowded
Online Store Management85946359414816%0.06Crowded, outpacing the catalogue
Memory & State85072732023012%0.05Crowded, slowing
Operations & Supply Chain81452843323140%0.18Crowded, slowing
Sales Automation81345055013650%0.28Crowded, outpacing the catalogue
Market Data78847343620037%0.19Crowded, slowing

Specific capabilities (level 4)

SEO Research is the largest capability segment with 1,073 products (735 vs 197 launches in the last two 90-day windows). 5 of the 20 largest are crowded and launching faster than the catalogue as a whole. The most concentrated among them is DeFi Protocols, where the leading project holds 87% of GitHub stars (HHI 0.77). The least concentrated is Multi-agent Coordination (top share 8%, HHI 0.03), where no single project dominates.

Largest capabilitiesDER

Largest capabilities
SegmentProductsWith repoNew 90dPrev 90dTop shareHHILabel
SEO Research1,07365973519737%0.18Crowded
PDF Processing1,06861871220039%0.21Crowded
Invoice Processing92727372214172%0.56Crowded, outpacing the catalogue
Agent Memory80469931222313%0.05Crowded, slowing
YouTube Content69737449512212%0.07Crowded, outpacing the catalogue
Code Quality59656435612221%0.08Crowded
Domains & DNS5733023959438%0.24Crowded, outpacing the catalogue
Consumer Messaging56131835510257%0.35Crowded
Cloudflare Platform54551328112042%0.22Crowded, slowing
Remote MCP Hosting51537227315144%0.23Crowded, slowing
Multi-agent Coordination489400122788%0.03Crowded, slowing
CRM Updates4722153108556%0.35Crowded
Semantic Search40934810311013%0.06Crowded, slowing
Headless Browsing3933401818712%0.06Crowded, slowing
Tax Compliance3741312945736%0.23Crowded, outpacing the catalogue
DeFi Protocols3602691509487%0.77Crowded, slowing
OCR3582322287729%0.13Crowded
Instruction Management348310735836%0.18Crowded, slowing
Webhooks & Triggers3301801938224%0.18Crowded, slowing
Campaign Management3241642284335%0.27Crowded, outpacing the catalogue

What this does not show: whether products in a capability are interchangeable. Two products tagged with the same capability may differ completely in quality, hosting model and maturity; the label only counts them.

Integrations

YouTube is the largest integration segment with 574 products (409 vs 98 launches in the last two 90-day windows). 7 of the 20 largest are crowded and launching faster than the catalogue as a whole. The most concentrated among them is Google Maps, where the leading project holds 72% of GitHub stars (HHI 0.56). The least concentrated is Azure (top share 15%, HHI 0.08), where no single project dominates.

Integrations mentioned by the most productsDER

Integrations mentioned by the most products
SegmentProductsWith repoNew 90dPrev 90dTop shareHHILabel
YouTube5742514099817%0.11Crowded, outpacing the catalogue
GitHub5523252459820%0.09Crowded, slowing
LinkedIn4201962877824%0.13Crowded
Solana3571582068661%0.45Crowded, slowing
PostgreSQL2871941406120%0.12Crowded, slowing
WhatsApp2851322004832%0.22Crowded, outpacing the catalogue
Shopify228851633853%0.35Crowded, outpacing the catalogue
Android2281551363215%0.08Crowded, outpacing the catalogue
Telegram2001141303341%0.25Crowded
Reddit1961061123638%0.22Crowded
Stripe18956827539%0.32Crowded, slowing
AWS162100433517%0.08Crowded, slowing
WordPress155821121970%0.52Crowded, outpacing the catalogue
Google Ads153741151847%0.44Crowded, outpacing the catalogue
Gmail12675632539%0.28Crowded, slowing
Azure12467381515%0.08Crowded, slowing
Kubernetes12092382621%0.11Crowded, slowing
Google Maps11949981072%0.56Crowded, outpacing the catalogue
Slack11472441638%0.21Crowded
MySQL11279452728%0.17Crowded, slowing

Integration counts come from product names and descriptions that mention the service. Official vendor-built servers and community wrappers count the same here; see the integration coverage report for demand against supply.

Recommendations

Editorial: the judgement of AgentSidekick, drawn from the figures above. Not a measurement, not investment advice.

  1. Editorial

    a new entrant in SEO Research (1,073 products) needs a clear differentiator; competing on "also does seo research" will not be visible.

  2. Editorial

    high concentration is a stronger barrier than size. In 139 segments one project already holds half the stars, so new projects there should integrate with the leader rather than replace it.

  3. Editorial

    the 120 thin segments are where a single well-documented product can become the reference; check demand before building, because thin can also mean unwanted.

Sources

Data behind this report

Cite this report

Most crowded categories in the agent ecosystem. AgentSidekick, 2026-11-21. https://www.agentsidekick.dev/reports/most-crowded-categories/

Figures change when the database is refreshed; cite the date shown.

Page data last updated .