UAI-1 is the current public message standard for structured AI-to-AI communication.
UAIX shield mark Public interoperability charter UAIX.org Universal Artificial Intelligence Exchange Public standards publication
Current UAI-1
  • Current version UAI-1
  • Plain-English definition Open message format for auditable AI-to-AI exchange
  • Current attribution Michael Joseph Kappel, MCP
  • Start here Get Started, Project Handoff, UAI-1, validator, roadmap, and changelog

Standards

UAIX Capability Ladder

L0-L6 capability ladder for static readers, structured fetchers, browser-assisted agents, tool agents, workflow agents, multi-agent runtimes, and audited systems.

  • Record UAIX-STND-2705
  • Path /en-us/standards/capability-ladder/
  • Use Canonical public record

Document status

Public standards page Published on UAIX as part of the current public standards record
Code
UAIX-STND-2705
Surface
Standards
Access
Public and linkable

How to use this page

Use this page to choose L0-L6 capability posture, fallbacks, and consent requirements before claiming support.

Proof path

Validator-backed proof path

Keep the public reading order tied to one evidence trail: profile, schema, example, validator result, and release record.

  1. 1Pick a message profile.Start with a published UAI-1 profile and the record family that matches the exchange you need to prove.
  2. 2Compare it with schemas and examples.Resolve the schema, registry entry, and one fixture before writing or mapping your candidate packet.
  3. 3Run validator evidence.Validate keyed, minified-keyed, or keyless JSON against the current public UAI-1 records.
  4. 4Attach the result to implementation or handoff records.Carry the exported result into Conformance Pack, implementation track, changelog, or Project Handoff evidence.

The UAIX capability ladder lets one standards surface degrade safely across very different clients. The level describes what a client can reliably read or do; it never grants permission by itself.

Level Client posture May do Must not imply Fallback
L0 Static or limited browser Read static HTML, text, Markdown, sitemap, llms, and small JSON. JavaScript, posting, auth, tools, memory, or coordination. Plain links and copyable text.
L1 Structured fetch Fetch predictable JSON, schemas, registry, examples, and manifests. Rendered UI parity or write-capable behavior. Low-token JSON and stable routes.
L2 Browser-assisted Inspect rendered UI and compare it to static evidence. Autonomous writes, auth, or tools. Static/JSON parity notes.
L3 Tool agent Call tools, authenticate, or post only with consent, idempotency, and evidence. Long workflow ownership or memory promotion. Blocker record and human review.
L4 Workflow agent Run external long tasks with correlation, task status, timeout, and final report. Unreviewed memory promotion or agent coordination. Status, blocker, and final-report packets.
L5 Multi-agent runtime Coordinate external agents while UAIX carries reviewed records and handoffs. UAIX-hosted execution or unverifiable trust claims. Unsigned reference handoff or verified receipt only when verifier evidence exists.
L6 Audited agent system Attach provenance, audit digest, consent, validator output, and public-safe receipts. Certification, endorsement, or trust without local policy. Public-safe audit digest and explicit exceptions.

Minimum rule

When capability is unknown, downgrade to L0/L1. Name the unsupported feature, return static or JSON evidence, and request human-mediated review before writes, auth, tools, memory, coordination, or audit claims.

Execution and trust boundary

Agent runtimes execute. UAIX records the reviewed communication, memory, trust, evidence, and handoff boundary.

  • UAIX standards records do not authorize execution, tool calls, repository writes, publication, payments, authentication, or durable memory promotion.
  • Local policy, server-side authorization, and human approval override envelope claims.
  • An unsigned envelope can be valid for review, routing, and evidence comparison, but it is not trusted identity.
  • Signed metadata is not trust by itself. A verifier may still report unknown, expired, revoked, disabled, locally untrusted, or otherwise unusable authority.
  • Reference examples for NeuralWikis, LocalEndpoint, Carcinus, MCP, A2A, and custom runtimes are consumer examples only. UAIX.org does not implement their runtime behavior.

Machine-readable companions