The Trust Standard for the Agentic Economy

We collect only the public artefacts you choose to submit: GitHub repository URL, MCP server endpoint, A2A protocol endpoint, or public website URL. No source code is mirrored. No production data is ingested. Scan results are stored encrypted in your account.

All data is stored in the European Union on Neon Postgres (Frankfurt) and Vercel Edge (Paris). Active customer data is retained for the duration of your subscription. After cancellation, audit data is deleted within 30 days unless you export it first.

Our 25-question checklist is aligned line-by-line with Regulation (EU) 2024/1689 Article 26 (deployer obligations). Each item maps to a concrete artefact: documentation, audit log, human oversight measure, or incident response procedure. We do not certify your compliance — we make your readiness state objectively measurable.

Service-level commitments for the Business and Enterprise tiers are defined in the signed DPA. Observed availability will be published monthly post-launch.

Yes. You can export and delete each audit from Settings. Account deletion is processed within the legal GDPR timeframe (Article 12: one month, extendable to three months if the request is complex). Email confirmation is sent once completed.

Every score is reproducible: weights, formulas, and risk bands are published in our public methodology pages. Authority dimension weighs 25% of the Agent-Readiness Score. The Trust Score uses 6 dimensions with documented LLM judges and embedding-based consistency tests.