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

AI Memory

Same-Codebase Agents

Stable logical identity, disjoint write scopes, task ownership, branch or worktree references, integration authority, and evidence for simultaneous code agents.

  • Record UAIX-MEMR-4024
  • Path /en-us/ai-memory/multi-agent-workloads/same-codebase/
  • Use Canonical public record

Document status

Public standards page Published on UAIX as part of the current public standards record
Code
UAIX-MEMR-4024
Surface
AI Memory
Access
Public and linkable

How to use this page

Use this page as part of the current AI Memory public record, then follow its linked standards pages for the next step.

For agents: apply the relevant guidance to the user's target site or repository. Do not copy this UAIX page, create a page about the guidance, or treat this page as runtime authority unless the user explicitly asks for that output.

Use the simple file-based single-agent package unless two or more agents truly need simultaneous ownership of one codebase. Same-codebase mode adds an explicit roster, disjoint write scopes, task ownership, branch or worktree references, integration authority, and acceptance evidence without replacing the local .uai/ startup suite.

Required ownership model

  1. Give every participant a stable logical ID and a unique registered name.
  2. Assign non-overlapping write scopes and one branch or worktree reference per code-writing agent.
  3. Bind every task to one roster owner and one acceptance-evidence rule.
  4. Name the integration authority, task-claim authority, shared-state authority, and conflict policy.
  5. Serialize any unavoidable shared scope through the named integration authority; do not rely on last-writer-wins.

Root and receiver records

.uai/swarm-profile.uai is the root workload manifest. Each .uai/agents/{logical-id}/receiver-brief.uai belongs to one logical agent and points back to that root. A per-agent brief may narrow work, but it cannot silently widen authority or override the root.

Completion evidence

Completion requires exact files or records changed, checks run or skipped, integration status, blockers, current proof, and one next action. A branch, task claim, or attempted write is not integration proof.

Open the Same-Codebase setup mode.

Support boundary

Agent runtimes execute. UAIX records portable configuration, logical identity, authority boundaries, routing references, reviewed evidence, and offline continuity. Runtime systems retain authentication, credential verification, leases, queues, transactions, tool execution, database writes, and delivery readback. A missing-authority request is review-pending only and applies no grants.

Schemas and validation