For API-first companies

Your API docs
make devs
build wrong.

RefLens compares your developer documentation against best-in-class references — Stripe, Twilio, Adyen — and identifies exactly where a human developer or AI agent would misunderstand your system and build the wrong thing.

Most doc tools measure completeness. RefLens measures correctness.

AUDIT RESULT 62 / 100
HIGH Evaluation vs Execution — docs describe endpoints, not protocol outcomes. Agent will call POST /consume without understanding that consumption is terminal, not recoverable.
MED Carrier data field blockReason documented as display-only. Agent will POST it expecting mutation — wrong model.
MED PASS / BLOCKED / DENIED outcomes listed without explaining recovery paths. Agent will treat BLOCKED as terminal.
73%
of API integration failures trace to documentation gaps — not code bugs
2.4×
more likely to fail if docs lack AI-agent readability markers
5 days
average time lost debugging integrations built on misunderstood docs

How it works

01

Input your docs

Paste your API documentation URL and OpenAPI spec. Or upload your spec file directly. RefLens accepts any REST API documentation format.

02

Pick your benchmark

Choose a reference API to compare against — Stripe (payments), Twilio (onboarding), Adyen (complex flows), Voucherify (loyalty), or a custom competitor. RefLens maps your domain concepts to the reference model.

03

Get your audit

RefLens scores your docs across orientation clarity, conceptual model coherence, first-call path, error recovery guidance, and AI-agent readiness. Each finding includes evidence, comparison to the benchmark, impact, and a prioritised fix recommendation.

What you get

Docs Scorecard

Composite score across 11 dimensions. Know at a glance where your documentation stands against best-in-class.

AI-Agent Readiness Score

Explicit assessment of whether a zero-knowledge AI coding agent can integrate without hallucinating. This is the score Stripe's docs pass. Yours probably don't.

Wrong Things Developers Build

The core output. RefLens identifies specific integration paths that will fail because your docs implied the wrong mental model — and shows exactly what needs to change.

Remediation Backlog

Prioritised list of documentation fixes, OpenAPI improvements, and example additions — each with expected impact, effort, and risk if ignored.

Developer Journey Simulation

Simulated first-call paths for both a human developer and an AI coding agent — where they get stuck, what they misunderstand, where they succeed.

Competitive Comparison

Side-by-side view of your docs against the chosen benchmark — concept coverage, example quality, error handling, and naming consistency.

Benchmarked against the best

RefLens evaluates your docs against these reference implementations — chosen because they define the standard in their respective domains.

Stripe
Payments — first-call journey, versioning, test mode
95
Twilio
Onboarding — quickstarts, SDKs, multi-language
91
Adyen
Complex flows — first-call guidance, sandbox, error recovery
88
Voucherify
Loyalty & redemption — object model, validation flow
82
Ventrata / OCTO
Booking standards — availability vs reservation vs redemption
79

Your docs are tested against whichever reference matches your domain. Or all five — if you want to know where you stand across the board.

Run your audit

Paste your documentation URL, pick a benchmark, and optionally add an OpenAPI spec. RefLens does the rest.

The main documentation page for your API
Optional. Add a direct OpenAPI/Swagger JSON or YAML link for deeper contract analysis. Leave blank for a docs-only audit.

If your docs make a developer build the wrong thing, you won't know until they tell you.

RefLens catches it before they start. Run your first audit and find out what's costing you integrations.