Agent Readiness — the competitive edge of being chosen by AI
We independently rate the Agent Readiness of companies and software, based on public facts and measured data.
View ARI Award 2026 SummerAre Japan's leading SaaS ready to be chosen by AI agents? We verified 200 SaaS services in the Japanese market against vendor primary sources and certified the 41 rated A or above. Methodology fully disclosed.
Every fact checked against vendor primary sources (2026-07-16).
AAA: 11, AA: 21, A: 9. Rules-based scoring, no LLM judgment.
19% of rated services provide an official MCP endpoint for AI agents.
Agent Readiness is not a single metric. We define it as five domains, each hardened progressively: public facts → first-party probes → real usage data from multiple AI agents. Every score carries a data-source label (public signal / probe / usage telemetry) — estimates are never presented as measurements.
| Domain | What it measures | Status |
|---|---|---|
| AI Access | Can AI connect? API, MCP, authentication | ✅ Published (2026 Summer) |
| AI Discoverability | Can AI find and understand product info and docs? | Partially measured — next issue |
| AI Execution | Can AI actually complete tasks? | In development |
| AI Trust | Is it safe to delegate? Permissions, audit, security | In development |
| AI Compatibility | How well does it work with ChatGPT, Claude, Gemini? | In development |
The evidence base behind our ratings — 11,000+ SaaS/MCP services, updated daily.
Our service ratings assess how ready a SaaS service is for AI agents. Using a rules-based methodology — no LLM judgment — we assign a letter rating based on MCP availability, API openness, documentation, and authentication clarity, supplemented by first-party connectivity probes.
We evaluate 11,000+ SaaS/MCP services across 23 categories; 2,257 services have been probe-tested by our own scanner via real JSON-RPC handshakes (as of 2026-07). Per-service success-rate figures are not published until independent third-party agent telemetry reaches sufficient scale.
The AEO (Agent Engine Optimization) Score used in this database measures primarily the AI Access and AI Discoverability domains of the Agent Readiness framework.
Six ways AI agents, SaaS companies, and integrators leverage KanseiLink ratings
Agents automatically discover the best SaaS for their purpose. Verified services are prioritized.
Pre-assess connection methods, verification status, and known pitfalls of target SaaS to minimize failure risk.
Get optimal integration patterns from 200 recipes (12 MCP-native with verified tool names). Build multi-service automation workflows.
SaaS companies compare their agent-readiness against industry benchmarks. Identify improvement areas.
Suggest specific actions to improve ratings: MCP adoption, API design, documentation.
Real-time tracking of API changes, MCP updates, and incidents. Protect agent workflows.
The scale and depth behind KanseiLink ratings.
Covering 11,000+ SaaS/MCP services across global and Japanese markets in 23 categories; 2,257 services probe-tested by our own scanner (as of 2026-07).
200 multi-service workflow patterns including 12 MCP-native recipes with verified tool names.
14 MCP tools including search, ratings, recipes, insights, and AEO reports.
Detailed analysis by category: accounting, HR, CRM, communication, project management, and more.
Live JSON-RPC handshake probes against hosted MCP endpoints — 3,475 successful out of 4,367 attempts by our own scanner (as of 2026-07). Third-party agent reports are being integrated progressively.
Ratings updated quarterly. Reflects changes in MCP adoption and new service additions.
Service-level ranking across 11,000+ global and Japan SaaS/MCP services (2,257 services probe-tested by our own scanner as of 2026-07). Click a service name to view its detailed profile. For company-level ratings of Japan's leading SaaS, see ARI Award 2026 Summer.
| # | Grade | Score | Service | Category | Status | MCP | Recipes | Success |
|---|---|---|---|---|---|---|---|---|
| 1 | AAA | 1.00 | Notion MCP | Verified | Official | 8 | High | |
| 2 | AAA | 0.90 | Slack MCP | Verified | Official | 35 | High | |
| 3 | AAA | 0.90 | freee MCP | Verified | Official | 19 | High | |
| 4 | AAA | 0.90 | Shopify Japan MCP | Verified | Official | 6 | High | |
| 5 | AAA | 0.90 | Backlog MCP | Verified | Official | 5 | High | |
| 6 | AAA | 0.85 | Brave Search | Verified | Official | 1 | N/A | |
| 7 | AAA | 0.85 | Playwright MCP | Verified | Official | 0 | N/A | |
| 8 | AAA | 0.85 | PostgreSQL MCP | Verified | Official | 0 | N/A | |
| 9 | AA | 0.80 | kintone MCP | Connectable | Official | 24 | Med | |
| 10 | AA | 0.80 | Chatwork MCP | Connectable | Official | 16 | Med | |
| 11 | AA | 0.80 | HubSpot Japan MCP | Connectable | Official | 5 | - | |
| 12 | AA | 0.80 | Money Forward Cloud MCP | Verified | Official | 4 | High | |
| 13 | AA | 0.80 | LINE Messaging MCP | Connectable | Official | 4 | - | |
| 14 | AA | 0.80 | freee HR | Connectable | Official | 4 | - | |
| 15 | AA | 0.80 | Sansan MCP | Connectable | Official | 2 | Med | |
| 16 | AA | 0.80 | Asana MCP | Connectable | Official | 1 | - | |
| 17 | AA | 0.80 | Tavily | Verified | Official | 0 | N/A | |
| 18 | AA | 0.80 | Firecrawl | Verified | Official | 0 | N/A | |
| 19 | AA | 0.75 | Google Drive | Connectable | Third-party | 3 | N/A | |
| 20 | AA | 0.75 | Google Analytics (GA4) | Connectable | API Only | 1 | N/A | |
| 21 | AA | 0.75 | Qdrant | Verified | Official | 0 | N/A | |
| 22 | A | 0.70 | SendGrid | Verified | Third-party | 2 | High | |
| 23 | A | 0.70 | Garoon MCP | Connectable | Official | 1 | Med | |
| 24 | A | 0.70 | DocuSign Japan | Connectable | Third-party | 1 | High | |
| 25 | A | 0.70 | Treasure Data MCP | Connectable | Official | 1 | Med | |
| 26 | A | 0.70 | 1Password Business | Connectable | Official | 1 | High | |
| 27 | A | 0.70 | Microsoft Teams MCP | Connectable | Official | 0 | High | |
| 28 | A | 0.70 | Twilio | Connectable | Third-party | 2 | N/A | |
| 29 | A | 0.70 | Amazon SES | Connectable | API Only | 0 | N/A | |
| 30 | A | 0.70 | AWS (Lambda / DynamoDB / RDS) | Connectable | Third-party | 0 | N/A |
Detailed rating reports for SaaS vendors, custom assessments for investors and procurement, free data access for agent builders. Ratings themselves are never for sale.
View solutionsHow ratings are calculated: rating criteria, data sources, and the five domains of Agent Readiness.
Read moreRatings are never for sale. Formula fully public. Conflicts disclosed.
Read moreGuide for checking and improving your service's rating. MCP adoption roadmap.
Learn more
For AI Agents: Access this data programmatically via npx @kansei-link/mcp-server.
Tools: search_services, generate_aeo_report, get_service_detail, find_combinations, get_recipe, submit_feedback.
JSON endpoint: /aeo-data.json
