Cube

Semantic layer and headless BI platform that sits between data warehouses and agent/app consumers, exposing curated business metrics via REST, GraphQL, and SQL APIs.

Evaluated Mar 07, 2026 (0d ago) v0.36.x
Homepage ↗ Repo ↗ Other cube semantic-layer headless-bi rest-api graphql sql-api data-warehouse metrics
⚙ Agent Friendliness
62
/ 100
Can an agent use this?
🔒 Security
84
/ 100
Is it safe for agents?
⚡ Reliability
80
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
--
Documentation
86
Error Messages
82
Auth Simplicity
82
Rate Limits
80

🔒 Security

TLS Enforcement
90
Auth Strength
85
Scope Granularity
80
Dep. Hygiene
82
Secret Handling
82

JWT-based security context enables row-level security per tenant without query modification. CUBEJS_API_SECRET must be rotated carefully — all issued tokens share the same secret.

⚡ Reliability

Uptime/SLA
82
Version Stability
80
Breaking Changes
78
Error Recovery
82
AF Security Reliability

Best When

Serving multiple agent consumers that need consistent, governed business metrics from a data warehouse without duplicating metric logic in each consumer.

Avoid When

You need real-time row-level data access — Cube's pre-aggregation and caching model trades freshness for query speed.

Use Cases

  • Agent queries consistent business metrics (MRR, DAU, conversion rate) through Cube's REST API without writing warehouse SQL directly
  • Build natural language to data pipelines where an LLM generates Cube query JSON from user questions and executes them via the REST API
  • Cache expensive warehouse aggregations in Cube's pre-aggregation layer so agent metric queries return in milliseconds instead of seconds
  • Expose a single GraphQL metrics API to multiple agent frontends so each agent gets the same governed metric definitions without data drift
  • Implement row-level security for multi-tenant agent deployments using Cube's security context to filter data by tenant ID at query time

Not For

  • Simple single-warehouse queries where adding a semantic layer abstraction creates unnecessary complexity and latency
  • Streaming or real-time event data requiring sub-100ms freshness — Cube's pre-aggregation model is optimized for batch refresh cycles
  • Teams without a data warehouse (Snowflake, BigQuery, Redshift, etc.) — Cube requires a supported warehouse as its data source

Interface

REST API
Yes
GraphQL
Yes
gRPC
No
MCP Server
No
SDK
Yes
Webhooks
No

Authentication

Methods: jwt api_key
OAuth: No Scopes: No

JWT tokens carry security context (tenant ID, user role) used for row-level security. API secret configurable via CUBEJS_API_SECRET env var. Cube Cloud adds token management UI.

Pricing

Model: open_source
Free tier: Yes
Requires CC: No

Apache 2.0 open source for self-hosted. Cube Cloud is the managed option with monitoring, CI/CD integration, and enterprise support.

Agent Metadata

Pagination
offset
Idempotent
Full
Retry Guidance
Documented

Known Gotchas

  • Pre-aggregation refresh lag means agent queries may return data that is hours old — agents must check the lastRefreshTime field to validate data freshness
  • Cube query JSON uses its own schema (measures, dimensions, filters) that differs from SQL — LLM agents generating Cube queries must be fine-tuned or prompted with the Cube schema format
  • Security context is embedded in the JWT token — agents that need to query as different tenants must generate separate tokens per tenant, not per query
  • GraphQL schema is auto-generated from cube definitions — adding/removing measures requires a schema reload before they appear in GraphQL introspection
  • REST /v1/load endpoint returns a 'Continue wait' 202 response for slow queries — agents must poll with the provided token until data is ready, implementing their own polling loop

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for Cube.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-07.

6470
Packages Evaluated
26150
Need Evaluation
173
Need Re-evaluation
Community Powered