ScoreFrameworkRankingPricingBlog GitHub

Kiwi.com

kiwi

87

Dominant

2 tools · Source: smithery · Scanned: April 18, 2026

Score breakdown

Findability

25/25

Can AI agents discover this tool? Measures naming quality, keyword overlap with queries, and registry presence.

Clarity

28/35

Can agents understand what this tool does? Measures description quality, action verbs, usage context ("Use this when..."), and return value documentation.

Precision

23/25

Is the interface precise? Measures input schema completeness, type definitions, required fields, enums, defaults, and parameter descriptions.

Efficiency

11/15

Is the tool token-efficient? Measures description length optimization, tool count, and naming conciseness.

Optimization status

Clarity

Add "Use this when..." context to your description so agents know when to invoke this tool

Expected impact: +3-5 pts

Efficiency

Keep total tool count under 15 and individual descriptions under 200 tokens

Expected impact: +1-3 pts

1 more improvement available

Unlock all fix suggestions, rewrite proposals, and priority recommendations.

Upgrade to Pro →

Estimated selection rate

~85%

At 87/100, AI agents select tools like this approximately 85% of the time when choosing between competing alternatives. This server is in the top tier for agent discoverability.

Based on 500-round selection simulation (Pearson r = 0.828). Methodology →

Trust Verification

1Spec Verified
📋Spec Verified✓ Verified
87/100Spec Score
RequirementScore ≥ 85 on ToolRank Spec Quality analysis
MethodAutomated static analysis of tool definitions, descriptions, parameters, error handling
RenewalRe-verified on every scan (weekly)
🎯Selection Verified○ Not yet
—Win Rate
RequirementSelection win rate ≥ 70% in 100-round LLM tournament
MethodClaude Sonnet selects best tool for realistic tasks vs. competing servers
RenewalRe-tested monthly; expires if win rate drops below 60%
⚡Runtime Verified○ Not yet
—Status
RequirementTask success rate ≥ 90%, error rate < 5%, avg latency < 2s
MethodEnd-to-end execution tests in sandboxed environment
RenewalContinuously monitored; revoked on sustained failure
View full audit log → ![Trust](https://mcp.toolrank.dev/api/trust-badge/f5901696-7f79-40c1-bcf1-32db9bdd972d.svg)

Score history

87
Apr 11
87
Apr 12
87
Apr 13
87
Apr 14
87
Apr 15
87
Apr 16
87
Apr 18

Tracked daily. 7 data points.

Similar servers

NarrateAI

92/100 · Dominant

Ssemble AI Clipping

92/100 · Dominant

pcx

92/100 · Dominant

Weather MCP Server

92/100 · Dominant

Domain Intel

92/100 · Dominant

licium

92/100 · Dominant
Score this server ← All rankings

Add ToolRank badge to this server's README

[![ToolRank](https://toolrank.dev/badge/dominant.svg)](https://toolrank.dev/ranking/kiwi)
Post on 𝕏 LinkedIn