Proposed public leaderboards
We are assembling fair, category-specific comparisons using the same scanner version and a disclosed observation window. No site is ranked publicly without owner opt-in.
8 proposed scans
Browser agents depend on reliable navigation, rendered semantics, and safe automation surfaces.
Inclusion: Public products whose primary offer is hosted browser automation, browser sessions, or agent browsing infrastructure.
8 proposed scans
Agents need to discover data models, APIs, authentication, limits, and safe write boundaries.
Inclusion: Managed database, data platform, or vector-search products with a public developer journey.
8 proposed scans
Voice-agent builders need clear model capabilities, latency claims, API paths, pricing, and consent boundaries.
Inclusion: Speech, transcription, voice generation, or conversational voice infrastructure with public APIs.
8 proposed scans
Agent-operated products need legible event contracts, diagnostics, documentation, and consequence-aware controls.
Inclusion: Product analytics, session analytics, error monitoring, logs, traces, or observability platforms.
8 proposed scans
These products should make agent concepts, tools, state, failure modes, and execution boundaries especially clear.
Inclusion: Frameworks and automation platforms explicitly supporting agent construction or orchestration.
8 proposed scans
Their visitors often arrive with agent-assisted intent and need machine-legible product, pricing, and deployment paths.
Inclusion: AI coding assistants, IDEs, and prompt-to-app development products with public onboarding.
8 proposed scans
Agents need fast API discovery, structured documentation, pricing clarity, and safe authentication guidance.
Inclusion: Developer platforms whose main user journey includes integrating a public API or SDK.
8 proposed scans
A focused cohort can surface practical patterns that smaller teams can actually adopt.
Inclusion: Independently operated SaaS products that opt in and disclose the primary journey being measured.
8 proposed scans
Commerce agents need explicit product data, transaction boundaries, confirmation steps, and support routes.
Inclusion: Payments, merchant-of-record, billing, checkout, or commerce API platforms.
Methodology 2026-08-27-v1
Candidate scans stay private while we verify category fit, scanner comparability, observation dates, and owner permission. Published rankings will include the denominator, timestamp, scanner version, evidence coverage, ties, and methodology changes.
Research without owner-ranking claims
The independent research index publishes task contracts, timestamps, evidence coverage, observable strengths, gaps, safe stopping points, and corrections. Numeric rankings and verified improvement claims remain opt-in.
Open the journey research