VerbVet
A CI-grade OpenAPI linter and migrator that scores every endpoint for HTTP-verb fitness and auto-suggests clean conversions to QUERY for bloated search routes.
API platform engineers maintaining large OpenAPI specifications at scale
- Heuristic scoring of every endpoint for GET, POST, and QUERY fitness based on payload size and complexity
- Auto-generated migration PRs that split bloated POST searches into clean GET + QUERY pairs
- Multi-SDK client regeneration with backward-compatibility shims and a flag-gated rollout plan
- Spec diffs and rollback-aware changelog reports for each migration step
The HTTP QUERY method was just formalized, and platform teams running 1,000+ endpoint specs need a migration path before their bloated POST searches become a standards-compliance liability.
QUERY method is real (RFC 10008) and enterprise OpenAPI specs are huge (50K-80K+ lines per Speakeasy), but the specific pain of migrating bloated POST searches to QUERY is a narrow, early-stage concern with little search-volume evidence.The HTTP QUERY Method RFC 10008 - IETF Datatracker ↗Why We Built Our Own OpenAPI Linter (And How It Compares) ↗
Spectral (open source, extensible custom rules) and Vacuum already cover broad OpenAPI linting; a custom Spectral rule could ship the QUERY-suggestion feature tomorrow, and Speakeasy just open-sourced another fast Go-based linter — incumbents are dense.GitHub - stoplightio/spectral: A flexible JSON/YAML linter ↗GitHub - daveshanley/vacuum: worlds fastest and most OpenAPI linter ↗
The dominant OpenAPI linters (Spectral, Vacuum, Speakeasy's new one) are all open source with no per-seat fees; monetization would require bundling into a broader governance platform à la Stoplight Enterprise or Scalar — a tough sell for a single-purpose CI tool.Spectral: Open Source API Description Linter | Stoplight ↗Enterprise-grade API Platform - scalar.com ↗
QUERY is freshly standardized and real-world adoption (server frameworks, API gateways, CDNs) is still unproven, so a standards-compliance migration tool may sit in the trough of disillusionment for years before demand materializes.IETF HTTP WG 2025-03 Interim - QUERY Method ↗
Trivial to build on top of mature OpenAPI 3 parsers (swagger-parser, kin-openapi); the heuristic for 'POST looks like a bloated search' plus a rewrite to QUERY is small-surface engineering — the hard part is adoption, not code.Why We Built Our Own OpenAPI Linter (And How It Compares) ↗