UAIX AI Instructions

This is the canonical AI-agent entry file for the UAIX workspace.

Metadata

Field	Value
Source site	uaix.org
Source URL	https://uaix.org/
Canonical AIWikis URL	https://aiwikis.org/uaix/uai-system/files/agents-md-e1faae8d/
Source reference	AGENTS.md
File type	md
Content category	uai-system
Last fetched	2026-05-15T00:23:56.0837262Z
Last changed	2026-05-09T18:10:22.7937253Z
Content hash	sha256:e1faae8dcb3e86d8c4bfc6d68e444c1daf9c91fb6ecab44d725986f9437901ad
Import status	changed
Raw source layer	data/sources/uaix/agents-md-e1faae8dcb3e.md
Normalized source layer	data/normalized/uaix/agents-md-e1faae8dcb3e.txt

Current File Content

Structure Preview

• UAIX AI Instructions
• Status
• Handoff Summary
• UAIX/Codex Handoff
• Workspace Coordination
• Loaded Context
• File Intake
• Required First Response
• Current State
• Next Steps
• Agent History
• Open Questions
• Deployment Rules
• Do Not Load By Default
• Do Not Change
• Markdown Maintenance Rules

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: 28981
• Preview characters: 11847

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: agents
project: "UAIX"
created: "2026-04-25"
updated: "2026-05-09"
status: active
version: 71
---

# UAIX AI Instructions

## Status

This is the canonical AI-agent entry file for the UAIX workspace.

Use `AGENTS.md` exactly as the front-door instruction file. Do not create `agent.md`, `Agents.md`, or another competing agent-instruction source. Tool-specific pointer files, such as `CLAUDE.md`, `GEMINI.md`, `.cursor/rules/uaix-handoff.mdc`, and `.github/copilot-instructions.md`, should remain lightweight redirects back to this file.

This file is intentionally concise. Detailed project truth belongs in root `readme.human` and the typed `.uai` files listed below.

## Handoff Summary

UAIX is the local WordPress Studio workspace for publishing UAIX.org, the public standards and publication site for UAI. UAI-1 is the current public exchange contract and evidence layer.

The repo-local handoff architecture is a specialized local path for Codex and other agents: load `AGENTS.md`, read `readme.human`, load typed `.uai` files, refresh dropped-file intake, and run targeted local checks. Direct UAIX.org pointing is for bootstrap, refresh, or public-source verification only; routine work should not depend on repeated live URL reading.

UAIX now uses AIWikis.org as long-term cold memory for bulky internal research, proposal history, roadmap background, and pre-slim handoff snapshots. Keep routine pickup focused on current UAIX hot context; open AIWikis raw memory only when the task explicitly needs original wording or older provenance.

UAIX.org is the UAI-1 source of truth. Protocol5.com is only a .NET plugin and NuGet distribution/backlink surface.

<!-- BEGIN UAIX-HANDOFF-MANAGED -->
## UAIX/Codex Handoff

Use the specialized local path for routine Codex and AI-agent work: read `AGENTS.md`, read `readme.human`, refresh active intake, load the typed `.uai` files, and run targeted local checks. Do not make recurring handoff work depend on live UAIX.org browsing.

The repo-local reusable Codex workflow is `.agents/skills/uaix-handoff/`. It exists to keep this handoff process repeatable while preserving `AGENTS.md` as the canonical front-door instruction file.
<!-- END UAIX-HANDOFF-MANAGED -->

## Workspace Coordination

This repo is part of a multisite Visual Studio Code workspace. Before broad work, read `[local path redacted]` when the human names any known site/domain, when a request spans multiple sites, or when the current shell directory and requested target differ.

Use the UAIX AI Memory Package Wizard at `https://uaix.org/en-us/tools/ai-memory-package-wizard/` for visible AI Memory setup guidance. Use the LLMWikis setup wizard at `https://llmwikis.org/tools/llm-wiki-setup-wizard/` when planning LLM Wiki structure, pointing long-memory/cold-memory routing to `https://AIWikis.org/` after source-site review.

Resolve the target site first. If the target is UAIX.org, continue with this file and the UAIX `.uai` bundle below. If the target is LLMWikis.org, AIWikis.org, JustAnIota.com, the short domain, or another registered site, switch to that site's `AGENTS.md` and load only that site's hot memory unless the task explicitly asks for cross-site source routing.

Before trusting a remembered absolute workspace path, verify that `workspace.uai` exists and that the selected registry root plus selected `AGENTS.md` path resolve. If the pointer is stale, locate the editor-workspace `workspace.uai` only enough to confirm the intended router, repair or report the stale path, and do not silently fall back to the current shell directory when the human named another target.

## Loaded Context

Load these files before broad planning or editing:

@uai[.uai/context.uai]
@uai[.uai/stack.uai]
@uai[.uai/architecture.uai]
@uai[.uai/public-surface.uai]
@uai[.uai/constraints.uai]
@uai[.uai/decisions.uai]
@uai[.uai/progress.uai]
@uai[.uai/operations.uai]
@uai[.uai/test-plan.uai]
@uai[.uai/style.uai]
@uai[.uai/handoff-pattern.uai]
@uai[.uai/file-handoff.uai]
@uai[.uai/intake-index.uai]

If any required `.uai` file is missing, unreadable, circular, contradictory, or resolves outside the repo, stop and report that before editing.

## File Intake

At the start of every broad AI work session, and whenever the newest prompt names an active intake bucket or says intake files were missed:

1. Inspect `agent-file-handoff/Content/`.
2. Inspect `agent-file-handoff/Improvement/`.
3. Ignore `agent-file-handoff/Archive/` unless the human explicitly names an archived file or moves it back into an active bucket.
4. Refresh `.uai/intake-index.uai` when a helper exists; this repo uses `scripts/Invoke-UaiFileIntake.ps1`.
5. If the prompt names `agent-file-handoff/`, `Content/`, `Improvement/`, a dropped file, or work that likely relates to active drops, treat intake as prompt-coupled: refresh the live buckets again even if startup intake already ran.
6. Summarize each `needs-agent-review` file before unrelated planning or edits.
7. For each safe active file, explicitly analyze its relationship to the current prompt before implementation.
8. Record one disposition per file: `apply-now`, `convert-into-roadmap-progress`, `archive-as-duplicate`, `defer-with-reason`, `ask-for-clarification`, or `block-as-unsafe-or-out-of-scope`.
9. For every safe, relevant file, extract at least one actionable slice and complete a real project-work action before archiving it. Work can be public copy, docs, tests, roadmap/progress, issue/evidence state, code, package metadata, or another accepted system surface.
10. Record the complete intake outcome: reviewed summary/disposition, hot-memory update or explicit no-change reason, long-memory/archive preservation when configured or explicit `not configured`, actual work completed, checks run or skipped, and blockers.
11. Use `defer-with-reason` only when no safe actionable slice can be promoted now; do not defer a safe, relevant report merely because it is broad, strategic, or future-facing.
12. Move processed source files to `Archive/` only after the disposition and completed work are recorded in durable state.
13. Never auto-execute blocked or untrusted files.

Indexing is not approval. Dropped files are local review inputs only, not public truth, release evidence, certification, endorsement, validator evidence, or permission to widen support claims. When a file is related to the current prompt, it must be pulled into the current work and understood against that task, not merely distributed into memory. Memory routing without a project-work result is a File Handoff failure unless every active file is unsafe, duplicate, out of scope, or truly blocked with a durable reason.

This workspace uses chat pickup, not CI pickup. Do not add CI jobs, scheduled jobs, repository automations, watchers, or background services for this handoff workflow; the active AI handles intake and targeted local checks inside the chat/work session.

## Required First Response

Before a broad change, reply with:

1. A 3-5 bullet summary of what UAIX is, what UAI-1 is, and what this repo publishes.
2. The active public theme, the public route shape, and the main machine-discovery assets.
3. The refreshed intake status, including the disposition for every `needs-agent-review` file.
4. The hard constraints from `.uai/constraints.uai`, especially support-claim boundaries, secrets/data boundaries, and destructive-operation limits.
5. The exact files or public routes expected to change for the current task.
6. The targeted checks expected to run, and any full release/package checks intentionally not running.

## Current State

- Active public theme: `wp-content/themes/uaix-authority-theme/`.
- Compatibility theme: `wp-content/themes/uaix-theme/`, still packaged and smoke-tested but not the production public surface.
- Public production routes use clean locale-prefixed paths under `/en-us/...`; enabled public languages are `en-US` and `zh-CN`.
- Long-form launch-review aliases redirect to canonical records: `/en-us/implementations/wordpress-publication-track/` to `/en-us/implementations/wordpress-reference/`, `/en-us/implementations/dotnet-bridge-track/` to `/en-us/implementations/dotnet-bridge/`, `/en-us/specification/agents-md-uai-linking-specification/` to `/en-us/specification/agents-md/`, and `/en-us/guides/agent-file-handoff/` to `/en-us/specification/agent-file-handoff/`.
- Main machine-discovery assets: `robots.txt`, `sitemap.xml`, `sitemap.html`, `/.well-known/uaix.json`, and `/.well-known/uai.json`.
- Public handoff and agentic-system surfaces include `/en-us/ai-memory/`, `/en-us/ai-memory/canonical-ai-memory/`, `/en-us/tools/ai-memory-package-wizard/`, `/en-us/specification/project-handoff/`, `/en-us/guides/agentic-harnesses-uai/`, `/en-us/guides/uaix-uai-1-agent-skill/`, `/en-us/guides/project-handoff-openai/`, `/en-us/guides/project-handoff-coding-agents/`, `/en-us/guides/project-handoff-context-budget/`, `/en-us/guides/project-handoff-knowledge-graphs/`, `/en-us/specification/agent-file-handoff/`, and `/en-us/specification/agents-md/`.
- UAIX.org publishes a reference UAIX / UAI-1 Agent Skill bundle for skills-compatible agents from `/en-us/guides/uaix-uai-1-agent-skill/` with the downloadable `uaix-uai-1-agent-skill-v0.1.0.zip` asset; `/en-us/guides/agentic-harnesses-uai/` now stays focused on the broader harness-versus-evidence strategy and points to the dedicated Agent Skill page. The bundle teaches agents how to create, validate, cite, and package UAI-1 evidence using UAIX public specification, validator, API routes, Adoption Kit, Conformance Pack, AI Memory, and Project Handoff records. Agent Skills handle reusable procedure loading; UAI-1 handles portable public evidence. This does not create an agent runtime, SDK, CLI, certification, endorsement, official adapter, hosted importer, automatic repository writer, automatic LLM Wiki sync, current capability-package profile, registry, authorization layer, or conformance support.
- Capability-package sidecars, package mapping annexes, package registries, package authorization fields, and capability-package conformance profiles are research-track roadmap ideas only. Current UAIX can evaluate optional evidence mappings for skill folders, plugin manifests, action groups, agent cards, MCP/OpenAPI capability surfaces, and subagent profiles, but it must not claim a current package profile, registry, authorization layer, official adapter, SDK, CLI, certification, endorsement, or conformance support.
- AI Memory Package Wizard setup paths now cover Base, Agent File Handoff, LLM Wiki, and the combined File Handoff plus LLM Wiki path. Generated artifacts preserve existing package state, keep LLM Wiki optional, link Canonical AI Memory layer guidance, capture project scope as Proof of Concept, Maintenance of Legacy System, or Greenfield Enterprise Software, ask whether setup is single-site `.uai` handoff or multisite `workspace.uai` coordination, require coordinator path/root/AGENTS.md verification before target memory loads, point LLM Wiki users to project coding standards, capture existing-system additive update mode, capture optional AI Dreaming Memory as a reviewed consolidation/replay policy over visible memory, capture multisite LLM Wiki source-routing strategy, capture optional LLM Wiki knowledge graph strategy/stable-ID/claim-provenance/validation/retrieval/export-boundary policies, and require the combined path to complete real project work plus hot/cold memory evidence before reporting completion. Copying a report into `.uai`, AIWikis, or an LLM Wiki is not project work by itself; a memory-only pass is a failed intake unless every active file is unsafe, duplicate, out of scope, or truly blocked with a durable reason.

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

AGENTS.md is the entry contract for agents. It tells an agent which context files to load, what authority boundaries apply, how file intake works, and which operations are out of bounds before broad edits start.

Structure

The file is structured around these visible headings: UAIX AI Instructions; Status; Handoff Summary; UAIX/Codex Handoff; Workspace Coordination; Loaded Context; File Intake; Required First Response. 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
• UAI system index
• Source provenance
• Global File Index
• Reports

Provenance And History

• Current observation: 2026-05-15T00:23:56.0837262Z
• Source origin: current-source-workspace
• Retrieval method: local-source-workspace
• Duplicate group: sfg-683 (primary)
• Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "UAIX AI Instructions",
    "source_site":  "uaix.org",
    "source_url":  "https://uaix.org/",
    "canonical_url":  "https://aiwikis.org/uaix/uai-system/files/agents-md-e1faae8d/",
    "source_reference":  "AGENTS.md",
    "file_type":  "md",
    "content_category":  "uai-system",
    "content_hash":  "sha256:e1faae8dcb3e86d8c4bfc6d68e444c1daf9c91fb6ecab44d725986f9437901ad",
    "last_fetched":  "2026-05-15T00:23:56.0837262Z",
    "last_changed":  "2026-05-09T18:10:22.7937253Z",
    "import_status":  "changed",
    "duplicate_group_id":  "sfg-683",
    "duplicate_role":  "primary",
    "related_files":  [

                      ],
    "generated_explanation":  true,
    "explanation_last_generated":  "2026-05-15T00:23:56.0837262Z"
}