Trust Observatory

Public trust cards on HUMΛN aggregate governed telemetry, benchmarks, and capability evidence. We reveal capability truth — we do not rank vendors.

Cards are sorted by last updated, not by score or vendor rank. Each card links to its methodology version and cites evidence windows explicitly.

Explore the Observatory

• Models — Published model trust cards with benchmark and governed telemetry evidence.
• Connectors — Connector reliability, setup friction, and policy compliance signals.
• Routes — Routing profiles including Cost Per Trusted Outcome where published.
• Agents — Agent trust cards for published marketplace and platform agents.
• Capabilities — Capability family cards with risk suitability and audit status filters.
• Methodology — How cards are built, what CPTO means, and the public/private boundary.

Related trust surfaces

• Agent authority — What an AI host can do, what it cannot do, and how you stay in control.
• What an AI host cannot do — A plaintext, enforceable list of prohibitions — engineering constraints, not marketing.
• Governance — Policy, provenance, and organizational control.
• Security — How we protect identity, secrets, and audit trails.

CLI access

The human trust command family exposes the Trust Observatory programmatically — public trust cards without login, plus org-private reports and webhooks when authenticated. See CLI trust docs on Console when available, or the public CLI guide mirrored under community docs.