← Leaderboard
9.0 L4

Clerk

Native Assessed · Docs reviewed · Mar 6, 2026 Confidence 0.64 Last evaluated Mar 6, 2026

Scores 9.0/10 overall. with execution at 9.0 and access readiness at 9.1. Payment: Self-serve billing, usage-based (MAU), credit card via web form. Governance: Per-key permissions, RBAC for organizations, audit logs, SOC 2 Type II, SCIM, SSO/SAML. Web accessibility: Dashboard is React SPA with decent accessibility.

Verify before you commit

Trust read first, source links second, build decision third.

Use this page to sanity-check Clerk quickly. We surface the evidence tier, freshness, and failure posture here, then put the official links where you can actually act on them, especially on mobile.

Evidence

Assessed

Docs reviewed · Mar 6, 2026

Freshness

Updated 2026-03-06T22:21:51.113+00:00

Mar 6, 2026

Failures

2

2 active failure modes

Score breakdown

Dimension Score Bar
Execution Score

Measures reliability, idempotency, error ergonomics, latency distribution, and schema stability.

9.0
Access Readiness Score

Measures how easily an agent can onboard, authenticate, and start using this service autonomously.

9.1
Aggregate AN Score

Composite score: 70% execution + 30% access readiness.

9.0

Autonomy breakdown

P1 Payment Autonomy
7.0
G1 Governance Readiness
8.0
W1 Web Agent Accessibility
7.0
Overall Autonomy 7.3/10
Ready for agent use

Active failure modes

  • UNKNOWN

    JWT verification clock skew

    Agent auth middleware rejects valid requests intermittently. Users appear randomly logged out. Hard to reproduce locally.

    Agent impact: Agent auth middleware rejects valid requests intermittently. Users appear randomly logged out. Hard to reproduce locally.

    Workaround: Add 10-second clock tolerance to JWT verification. Sync container clocks via NTP. Use Clerk's backend SDK which handles this automatically.

    occasional
  • UNKNOWN

    Webhook signature verification timing

    Agent loses webhook events when processing queue has any latency. No retry from Clerk side — event is permanently lost.

    Agent impact: Agent loses webhook events when processing queue has any latency. No retry from Clerk side — event is permanently lost.

    Workaround: Process webhook verification immediately on receipt, queue the payload for async processing after verification. Increase tolerance window if possible.

    occasional

Reviews

Published review summaries with trust provenance attached to each card.

How are reviews sourced?

Docs-backed Built from public docs and product materials.

Test-backed Backed by guided testing or evaluator-run checks.

Runtime-verified Verified from authenticated runtime evidence.

Clerk: Auth & Security Model

Test-backed

For Humans 1. Sign up at clerk.com(https://clerk.com) and create a new application. 2. Navigate to the API Keys section in the Clerk Dashboard. 3. Copy the CLERK_SECRET_KEY and CLERK_PUBLISHABLE_KEY. 4.

Rhumb editorial team Mar 10, 2026

Clerk — Agent-Native Service Guide

Test-backed

Clerk is a high-velocity authentication and user management platform designed to decouple identity from core application logic. For agents, Clerk serves as the programmatic gatekeeper for user lifecycle management, organization multi-tenancy, and session validation. It is particularly valuable for "Agent-in-the-Loop" systems where an autonomous... Reviewed from official documentation.

Rhumb editorial team Mar 10, 2026

Clerk: Error Handling & Reliability

Test-backed

Value :--- 80ms 180ms 300ms Variable 99.9% --- Idempotency: Clerk does not support a global Idempotency-Key header.

Rhumb editorial team Mar 10, 2026

Clerk: API Design & Integration

Test-backed

REST API Clerk provides a comprehensive Backend API accessible at https://api.clerk.com/v1. This is the primary interface for agents. Authentication is handled via a Secret Key passed in the Authorization: Bearer <CLERK_SECRET_KEY> header.

Rhumb editorial team Mar 10, 2026

Clerk: Documentation & Developer Experience

Test-backed

Clerk is a high-velocity authentication and user management platform designed to decouple identity from core application logic.

Rhumb editorial team Mar 10, 2026

Use in your agent

mcp
get_score ("clerk")
● Clerk 9.0 L4 Native
exec: 9.0 · access: 9.1

Trust shortcuts

This score is documentation-derived. Treat it as a docs-based evaluation of API design, auth, error handling, and documentation quality.

Read how the score works, how disputes are handled, and how Rhumb scored itself before launch.

Overall tier

L4 Native

9.0 / 10.0

Alternatives

No alternatives captured yet.