ToolPicker
// methodology

The ToolPicker Score

A transparent, auditable rubric — 100 points across six pillars. It measures how ready, transparent and well-built a tool is for AI agents to actually use. Every criterion is objective and public, so you can see exactly why a tool ranks where it does.

35
Agent-readiness
agent registration — auth.md / OAuth / API key (9) · public API (5) · OpenAPI (9) · MCP server (9) · llms.txt (3)
15
Reliability & performance
actively maintained (8) · production-ready (4) · updated recently (3)
13
Pricing
itemized public pricing (6) · free tier (4) · free-tier generosity (3)
12
Docs & DX
docs quality (8) · public API (2) · quickstart (2)
12
Security & compliance
SOC 2 (6) · ISO 27001 (2) · GDPR (2) · security / trust page (2)
10
Openness
open source (6) · open / un-gated API (4)
3
Company maturity
company size (2) · years operating (1)

No black box. Each tool page breaks the score down pillar by pillar — agent-readiness, pricing, security, docs — so you can see exactly which signals a tool has and audit the ranking yourself.

What it is not. It doesn't claim which tool is "best" at its core job — that's subjective and context-dependent. It measures how well a tool is set up for agents. Where we can run a real benchmark (e.g. live enrichment accuracy), we show it as a separate, empirical signal next to the score.

Company maturity is capped at 3. A sharp agent-native startup still outranks a big legacy vendor — maturity is a tie-breaker for stability, not a thumb on the scale for incumbents.