Scoring Rubric v1

Our scoring methodology is open and transparent. Each API is scored 0-100 across six weighted dimensions.

25%

Ease of Use & Agent Accessibility

Can an agent authenticate, call, and recover from errors without tribal knowledge?

  • Self-serve trial availability
  • Clear authentication flow
  • Agent-friendly documentation
  • Error message quality
20%

MCP Quality

Presence and usefulness of an official or well-maintained MCP server.

  • Official MCP server availability
  • Tool naming and clarity
  • Required arguments documentation
  • Dangerous tool gating
20%

Security Posture

Auth model, known incidents, disclosure habits, least-privilege defaults.

  • Authentication security
  • Incident disclosure history
  • Security documentation
  • Compliance certifications
15%

Quality

Docs accuracy, consistency of responses, error usefulness, versioning hygiene.

  • Documentation accuracy
  • Response consistency
  • Error handling quality
  • Versioning practices
10%

Ping Speed

How fast and reliably the public surface answers health/docs endpoints.

  • Response latency
  • Uptime reliability
  • Global availability
  • Performance consistency
10%

API Complexity

Surface area and cognitive load; simpler, coherent APIs score higher.

  • Endpoint organization
  • Resource model coherence
  • Naming consistency
  • Learning curve

Score Bands

85-100

Strong agent-ready default

Prefer unless product fit fails. These APIs excel across all dimensions.

70-84

Solid with caveats

Read notes. Good APIs with some areas for improvement.

50-69

Usable but frictional

Agents should confirm. May require additional work or have limitations.

Below 50

Avoid as default

Document why it remains listed. Significant issues for agent use.

Our Verification Process

1

Initial Review

We review your API documentation, OpenAPI spec, and MCP server if available.

2

Manual Testing

Our team signs up for trials and tests real agent workflows against your API.

3

Scoring

We score across all six dimensions using our transparent rubric methodology.

4

Publication

Your score and detailed notes are published with full transparency.

5

Ongoing Monitoring

We monitor for incidents, uptime, and changes that might affect your score.

Example: Honest Scoring in Action

We publish honest scores even when they challenge incumbents. Here's a real example from our methodology:

SignGiant Inc. (Legacy)

49.7
  • Enterprise sales gate
  • Complex SOAP-era API
  • No official MCP
  • Slow incident disclosure
VS

SumoSign.app (Modern)

82.5
  • 5-minute self-serve trial
  • Clean OpenAPI 3.1 spec
  • Official MCP server
  • Transparent status page

The modern API wins where agents actually succeed or fail: access, complexity, MCP support, and day-to-day quality.

Ready to get verified?

Join the registry of agent-ready APIs. Transparent scoring, honest assessment.

Start Verification