UAIX Decisions Log

UAIX Decisions Log is a uai-system source file exposed by AIWikis.org with provenance, current content, and generated explanation.

Metadata

Field	Value
Source site	uaix.org
Source URL	https://uaix.org/
Canonical AIWikis URL	https://aiwikis.org/uaix/files/raw-system-archives-uaix-internal-memory-reorg-2026-05-01-uai-decisions-b1600149/
Source reference	raw/system-archives/uaix/internal-memory-reorg/2026-05-01/.uai/decisions.uai
File type	uai
Content category	uai-system
Last fetched	2026-06-22T01:56:21.9510185Z
Last changed	2026-05-01T17:53:40.4135173Z
Content hash	sha256:b1600149dfd19795f1b754b0cc5e22b863aae551f3480b92d892170e2fd34459
Import status	unchanged
Raw source layer	data/sources/uaix/raw-system-archives-uaix-internal-memory-reorg-2026-05-01-uai-decisions-uai-b1600149dfd1.uai
Normalized source layer	data/normalized/uaix/raw-system-archives-uaix-internal-memory-reorg-2026-05-01-uai-decisions-uai-b1600149dfd1.txt

Current File Content

Structure Preview

• UAIX Decisions Log
• [2026-05-01] Keep coding-agent integrations as thin pointers to one Project Handoff bundle
• [2026-05-01] Add a repo-local Codex skill for UAIX handoff pickup
• [2026-05-01] Standardize the repo-local Codex handoff path on AGENTS.md
• [2026-05-01] Position Project Handoff as portable project memory for OpenAI-compatible workflows
• [2026-05-01] Make the AI Memory Package Wizard a staged review flow
• [2026-04-30] Publish UAI-1 comparison guidance on existing routed pages
• [2026-04-30] Populate AI Memory operating protocols in the Package Wizard
• [2026-04-30] Add a live desktop artifact preview to the AI Memory Package Wizard
• [2026-04-30] Generate operating system profiles from the AI Memory Package Wizard
• [2026-04-30] Use AIWikis outcome documentation for explicit archive consolidation feedback
• [2026-04-30] Make receiver startup briefs a first-class AI Memory Wizard output
• [2026-04-30] Add one-file startup packets for receiver operations
• [2026-04-30] Harden generated LLM Wiki memory-plan files as local planning artifacts
• [2026-04-29] Add AI Memory Package Wizard as guided canonical bundle UI
• [2026-04-29] Require transfer evidence and history for AIWikis archive consolidation
• [2026-04-29] Publish the AIWikis/LLMWikis/UAIX relationship audit as a bounded role record
• [2026-04-28] Treat AIWikis package dogfood as package-scaffold evidence
• [2026-04-28] Model AI Memory as a canonical bundle taxonomy
• [2026-04-28] Make AI Memory the broad public handoff phrase
• [2026-04-27] Archive already-dispositioned intake files
• [2026-04-27] Split file intake into Content and Improvement buckets
• [2026-04-27] Add readme.human as the human-facing handoff companion
• [2026-04-27] Make file intake chat-start local behavior

Raw Version

This public page shows a bounded preview of a large source file. The complete source remains in the raw and normalized source layers named in metadata, with the SHA-256 hash above for verification.

• Source characters: 37413
• Preview characters: 11808

Local absolute paths are redacted in this public view. The source hash and source-side raw layer are based on the unredacted source file.

---
uaix: "1.0"
type: decisions
title: "UAIX Decisions Log"
created: "2026-04-25"
updated: "2026-05-01"
status: active
author: "UAIX maintainers"
version: 22
---

# UAIX Decisions Log

## [2026-05-01] Keep coding-agent integrations as thin pointers to one Project Handoff bundle

**Decision:** UAIX now publishes `/en-us/guides/project-handoff-coding-agents/` for cross-tool coding-agent pickup and keeps local tool-specific files as thin pointers to `AGENTS.md`, `readme.human`, and `.uai` records. The local pointers are `CLAUDE.md`, `GEMINI.md`, `.cursor/rules/uaix-handoff.mdc`, and `.github/copilot-instructions.md`.

**Why:** The Codex, Claude Code, Cursor, GitHub Copilot, Gemini Code Assist, and direct-pointing reports all point to the same practical need: a future agent should load one repo-local handoff bundle instead of rediscovering project truth from private chat, public pages, or duplicated tool instructions.

**Effect:** The public guide, theme seed route, guide index, navigation, well-known manifests, artifact catalog, static discovery test, launch-surface audits, production response check, and focused AI Memory regression test now know the coding-agents guide route. Local tool pointers exist for the most likely coding-agent surfaces while preserving `AGENTS.md` as the front door.

**Boundary:** These are compatibility patterns and local pointer files, not official UAIX adapters, SDKs, CLIs, hosted import validators, automatic repository writers, vendor certifications, UAI-1 conformance evidence, validator-backed handoff compliance, or endorsement claims.

## [2026-05-01] Add a repo-local Codex skill for UAIX handoff pickup

**Decision:** UAIX keeps `AGENTS.md` as the canonical agent instruction front door and adds `.agents/skills/uaix-handoff/` as a local Codex skill for repeating the handoff load, active intake review, context loading, and targeted-check workflow.

**Why:** The active Codex, Cursor, Claude Code, Gemini Code Assist, Copilot, and direct-pointing reports converged on the same operational lesson: recurring agent work should use local, versioned, concise handoff artifacts instead of repeated live URL reading or tool-specific duplicate instruction files. A local skill gives Codex a reusable workflow without bloating `AGENTS.md` or making public product claims.

**Effect:** Future Codex sessions can use the skill to follow the UAIX pickup sequence: read `AGENTS.md`, read `readme.human`, refresh `agent-file-handoff/` intake, load `.uai` records, disposition active files, run targeted checks, and report skipped full release checks. Tool-specific files stay thin pointers; `CLAUDE.md` now imports `AGENTS.md` rather than duplicating project truth.

**Boundary:** The skill is repo-local workflow aid only. It is not a public UAIX adapter, SDK, CLI, hosted importer, automatic repository writer, certification, endorsement, UAI-1 conformance evidence, validator-backed status, CI pickup, or background intake automation.

## [2026-05-01] Standardize the repo-local Codex handoff path on AGENTS.md

**Decision:** UAIX repo-local handoff work standardizes on root `AGENTS.md` as the canonical front-door agent instruction file, with `readme.human` as the human briefing and typed `.uai` files as the detailed local context. Do not create `agent.md`, `Agents.md`, or another competing instruction source; compatibility files such as `CLAUDE.md` should remain short pointers back to `AGENTS.md`.

**Why:** The active handoff reports both recommend a specialized local workflow over repeated live URL reading. Local Markdown, JSON, YAML, and lintable scripts are faster, more reproducible, easier to review, and safer for Codex-style work than asking every session to browse UAIX pages and reinterpret the pattern. Direct UAIX.org pointing remains useful for bootstrap, refresh, and public-source verification.

**Effect:** The root handoff bundle now treats `AGENTS.md` as the concise operational entry point, keeps detailed context in `.uai` records, preserves `agent-file-handoff/` as the active dropped-file intake path, and adds local handoff linting for required files, required sections, `@uai[]` references, intake-index completeness, blocked extensions, and obvious secret patterns. Handoff pickup and linting are chat-session responsibilities, not CI jobs.

**Boundary:** Repo-local `AGENTS.md`, `readme.human`, `.uai`, and `agent-file-handoff/` files are not UAIX certification, endorsement, UAI-1 conformance, or validator-backed evidence. Hosted `.uai` import validation, automatic repository writes, SDK, CLI, certification, endorsement, official adapter support, automatic live-web refresh, CI pickup, and background intake automation remain out of current support unless explicit public evidence and human approval are added later.

## [2026-05-01] Position Project Handoff as portable project memory for OpenAI-compatible workflows

**Decision:** UAIX Project Handoff should be positioned as the portable project-memory, governance, and review layer that agent runtimes consume before work and update after work. It should not be positioned as an OpenAI replacement, agent runtime, scheduler, orchestrator, competing tool-calling framework, private memory store, or chat transcript archive.

**Why:** The active OpenAI orchestration reports converged on the same boundary: OpenAI and similar platforms are strong at runtime orchestration, tools, handoffs, approvals, sessions, traces, Skills, MCP, and execution loops. UAIX Project Handoff is strongest as durable repo-local transfer context: `AGENTS.md`, `readme.human`, typed `.uai` records, constraints, decisions, progress, verification plans, and reviewable human briefing.

**Effect:** `/en-us/specification/project-handoff/` now leads with "Project Handoff is portable project memory for humans and AI agents," includes a "What this is / What this is not" section, compares OpenAI runtime orchestration with Project Handoff, and repeats the boundary "OpenAI runs the agents. Project Handoff preserves the project memory." UAIX also publishes `/en-us/guides/project-handoff-openai/` for OpenAI Agents and Codex workflows, and the discovery manifests/tests know the new route.

**Boundary:** `.uai` JSON Schemas, Project Handoff validator, OpenAI adapter, trace-to-handoff exporter, example repos, and conformance tests remain planned or recommended work until public fixtures, validation behavior, release evidence, and ownership exist. This decision does not create hosted upload/import validation, automatic repository writes, automatic LLM Wiki sync, SDK, CLI, certification, endorsement, official OpenAI adapter support, or a claim that Project Handoff replaces OpenAI, Codex, Skills, MCP, or runtime orchestration.

## [2026-05-01] Make the AI Memory Package Wizard a staged review flow

**Decision:** The AI Memory Package Wizard should stay on the existing canonical route and become a six-step in-page flow: preset, basics, operations, protocols, receiver, and review/export. The current route now owns progress navigation, page-level validation, local browser draft restore, sensitive/high-risk review gate auto-enabling, optional LLM Wiki planning inside the receiver step, lazy review previews, and export actions only on the review step.

**Why:** The active wizard redesign reports called for progressive disclosure, a true review stage, clearer validation, state persistence, fewer visible simultaneous controls, and a stronger "generate after review" boundary. Keeping the work on the current route preserves public discovery, sitemap, docs, tests, and existing canonical bundle generation while reducing first-run complexity.

**Effect:** The shortcode markup, JavaScript controller, styling, focused AI Memory regression test, public route copy, and handoff state now describe and verify the staged flow. Refresh cadence and file-deck scope remain available as advanced review/export controls, while LLM Wiki remains optional package-tailoring guidance.

**Boundary:** This is local browser package planning over canonical starter bundles. It does not create hosted upload/import validation, automatic repository writes, automatic LLM Wiki sync, provider/API-key execution tests, MCP/server discovery, SDK, CLI, certification, endorsement, a WordPress admin installer, or automatic site writes.

## [2026-04-30] Publish UAI-1 comparison guidance on existing routed pages

**Decision:** UAIX should satisfy the initial "what is UAI-1", UAI-1 vs MCP, and UAI-1 vs A2A onboarding need on existing public routes rather than creating new comparison pages before locale, sitemap, and release evidence are ready.

**Why:** The roadmap called for source-backed comparison-led adoption content, but adding new public routes would create a larger localization and discovery surface. The existing Get Started, UAI-1, and Standards Fit pages already carry the right authority and navigation context, so deepening those pages finishes the user-facing need with less launch risk.

**Effect:** Get Started now points readers to Standards Fit for the full chooser; UAI-1 links to the fuller decision guide; Standards Fit now includes a quick chooser, UAI-1 vs MCP, UAI-1 vs A2A, decision questions, and bridge-evidence boundaries; route audits assert the new comparison markers; and roadmap/progress state treats the initial comparison/onboarding chunk as current public work.

**Boundary:** This is explanatory and evidence-routing guidance only. It does not create MCP or A2A runtime support, formal bridge profiles, SDKs, CLIs, certification, compliance approval, endorsement, or a claim that UAI-1 replaces adjacent protocols.

## [2026-04-30] Populate AI Memory operating protocols in the Package Wizard

**Decision:** The AI Memory Package Wizard now asks for source-authority policy, memory-update policy, evidence ledger path, conflict-resolution policy, change-risk level, memory update targets, and rollback strategy, then carries those answers into the generated overlay, package model, `UAI_MEMORY_SYSTEM_PROFILE.md`, `UAI_MEMORY_RECEIVER_BRIEF.md`, `UAI_MEMORY_STARTUP_PACKET.md`, and optional `LLM_WIKI_MEMORY_PLAN.md`.

**Why:** A memory package that only names files and generic tests still leaves too much interpretation for the receiving actor. The wizard should produce a usable memory operating layer that tells the next human, team, or AI which records outrank generated memory, when durable memory updates, where evidence is recorded, how contradictions are settled, when risk escalates, and how rollback is handled.

**Effect:** The wizard UI has a protocol and evidence step; generated system profiles include a memory system protocol map, source-authority and conflict protocol, evidence ledger protocol, and risk/rollback protocol; startup packets and receiver briefs summarize those choices; optional LLM Wiki plans inherit source-authority/conflict/evidence rules for stale or conflicting wiki memory; homepage, roadmap/docs, machine-roadmap state, and focused regression coverage now reflect the protocol expansion.

**Boundary:** These are generated local planning and handoff files beside canonical starter bundles. They do not create hosted upload/import validation, automatic repository writes, automatic LLM Wiki writes or sync, SDK, CLI, certification, endorsement, WordPress admin installation, or UAI-1 conformance evidence.

## [2026-04-30] Add a live desktop artifact preview to the AI Memory Package Wizard

**Decision:** The AI Memory Package Wizard now includes a sticky desktop build panel with live generated-file metrics, artifact tabs, and a live preview for `UAI_MEMORY_STARTUP_PACKET.md`, `UAI_MEMORY_SYSTEM_PROFILE.md`, `UAI_MEMORY_RECEIVER_BRIEF.md`, `UAI_MEMORY_WIZARD_OVERLAY.json`, and the optional `LLM_WIKI_MEMORY_PLAN.md`.

Why This File Exists

This is a UAI AI Memory handoff file from uaix.org. It is shown here because AIWikis.org is demonstrating the real source files that make the UAIX / LLM Wiki memory system work, not only summarizing those systems after the fact.

Role

This .uai file is a compact decisions packet. It keeps one kind of durable project truth separate from the rest of the archive so an agent can load the topic it needs without pulling the whole project history into prompt context.

Structure

The file is structured around these visible headings: UAIX Decisions Log; [2026-05-01] Keep coding-agent integrations as thin pointers to one Project Handoff bundle; [2026-05-01] Add a repo-local Codex skill for UAIX handoff pickup; [2026-05-01] Standardize the repo-local Codex handoff path on AGENTS.md; [2026-05-01] Position Project Handoff as portable project memory for OpenAI-compatible workflows; [2026-05-01] Make the AI Memory Package Wizard a staged review flow; [2026-04-30] Publish UAI-1 comparison guidance on existing routed pages; [2026-04-30] Populate AI Memory operating protocols in the Package Wizard. Those headings are retrieval anchors: a crawler or LLM can decide whether the file is relevant before reading every line.

Prompt-Size And Retrieval Benefit

Keeping this material in a separate file reduces prompt pressure because an agent can load this exact unit only when its role, source site, category, or hash is relevant. The surrounding index pages point to it, while this page preserves the full content for audit and exact recall.

How To Use It

• Humans should read the metadata first, then inspect the raw content when they need exact wording or provenance.
• LLMs and agents should use the source site, category, hash, headings, and related files to decide whether this file belongs in the active prompt.
• Crawlers should treat the AIWikis page as transparent evidence and follow the source URL/source reference for authority boundaries.
• Future maintainers should regenerate this page whenever the source hash changes, then review the explanation if the role or structure changed.

Update Requirements

When this source file changes, update the raw source layer, normalized source layer, hash history, this rendered page, generated explanation, source-file inventory, changed-files report, and any source-section index that links to it.

Related Pages

• Source overview
• Site file index
• Site report index
• UAI system index
• Source provenance
• Site directory
• Organization reports

Provenance And History

• Current observation: 2026-06-22T01:56:21.9510185Z
• Source origin: current-source-workspace
• Retrieval method: local-source-workspace
• Duplicate group: sfg-850 (primary)
• Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "UAIX Decisions Log",
    "source_site":  "uaix.org",
    "source_url":  "https://uaix.org/",
    "canonical_url":  "https://aiwikis.org/uaix/files/raw-system-archives-uaix-internal-memory-reorg-2026-05-01-uai-decisions-b1600149/",
    "source_reference":  "raw/system-archives/uaix/internal-memory-reorg/2026-05-01/.uai/decisions.uai",
    "file_type":  "uai",
    "content_category":  "uai-system",
    "content_hash":  "sha256:b1600149dfd19795f1b754b0cc5e22b863aae551f3480b92d892170e2fd34459",
    "last_fetched":  "2026-06-22T01:56:21.9510185Z",
    "last_changed":  "2026-05-01T17:53:40.4135173Z",
    "import_status":  "unchanged",
    "duplicate_group_id":  "sfg-850",
    "duplicate_role":  "primary",
    "related_files":  [

                      ],
    "generated_explanation":  true,
    "explanation_last_generated":  "2026-06-22T01:56:21.9510185Z"
}