UAIX Org And Agent Skills Integration Report

This report focuses strictly on **UAIX.org** and **Agent Skills**. It does **not** analyze or rely on **AUiX**, which is unrelated to UAIX.org. UAIX currently positions **UAI-1** as the public, portable exchange and e...

Metadata

Field	Value
Source site	uaix.org
Source URL	https://uaix.org/
Canonical AIWikis URL	https://aiwikis.org/uaix/files/raw-system-archives-uaix-agent-file-handoff-retired-source-archive-2026-1850145c/
Source reference	raw/system-archives/uaix/agent-file-handoff/retired-source-archive-2026-06-13/2026-05-06/Improvement/agent-skills-integration/UAIX.org and Agent Skills Integration Report 2.md
File type	md
Content category	memory-file
Last fetched	2026-06-22T01:56:21.9510185Z
Last changed	2026-05-06T22:33:37.3508317Z
Content hash	sha256:1850145c5d856814f01ad748f1ccf22f41c42485d625a582c696f5a15b792c9f
Import status	unchanged
Raw source layer	data/sources/uaix/raw-system-archives-uaix-agent-file-handoff-retired-source-archive-2026-06-13-2026-05-06-improve-1850145c5d85.md
Normalized source layer	data/normalized/uaix/raw-system-archives-uaix-agent-file-handoff-retired-source-archive-2026-06-13-2026-05-06-improve-1850145c5d85.txt

Current File Content

Structure Preview

• UAIX.org and Agent Skills Integration Report
• Executive summary
• Source basis and analytical fit
• Data flow and reference architecture
• Proposed UAIX Agent Skill bundle
• scripts/fetch_catalog.py
• Pseudocode: resolve live UAIX machine surface before route-specific work.
• scripts/validate_uai.py
• Pseudocode: validate a candidate packet and return a structured result.
• scripts/package_first_proof.py
• Pseudocode: assemble a minimal reusable proof packet from UAIX public sources.
• UAIX.org copy, rollout plan, and evaluation
• Using Agent Skills with UAIX
• What Agent Skills do
• What UAIX does
• How they work together
• What not to claim
• Start here
• Risks, controls, assumptions, and conclusion

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: 38620
• Preview characters: 11937

# UAIX.org and Agent Skills Integration Report

## Executive summary

This report focuses strictly on **UAIX.org** and **Agent Skills**. It does **not** analyze or rely on **AUiX**, which is unrelated to UAIX.org. UAIX currently positions **UAI-1** as the public, portable exchange and evidence layer for AI-to-AI work, with published schemas, a registry, examples, a validator, an API surface, an Adoption Kit, a Conformance Pack, AI Memory, Project Handoff, implementation tracks, and a public roadmap. The live catalog currently identifies **UAI-1** as the current release, with `uai_version` **1.0**, release date **2026-04-22**, and counts of **6 schemas**, **6 registry entries**, **6 examples**, **2 transport bindings**, **5 trust channels**, **4 conformance levels**, and **11 error codes**. citeturn3view0turn11view0turn35view0turn35view1

The strongest integration model is **layered, not overlapping**. **Agent Skills** should supply the reusable operating procedure: discovery, activation, instructions, bundled scripts, references, and evaluations. **UAIX / UAI-1** should supply the **portable record**, the **validator-backed evidence**, the **support-boundary language**, and the **handoff artifacts** that survive beyond one agent session or runtime. UAIX itself explicitly says it is the public evidence and handoff layer, not the runtime, and its roadmap warns not to describe UAIX as an agent runtime, scheduler, SDK, CLI, official adapter set, or certification body. citeturn5view0turn21view0turn30view1turn30view6

That positioning is also consistent with Agent Skills. The Agent Skills format defines a skill as a folder containing `SKILL.md` plus optional `scripts/`, `references/`, and `assets/`; agents discover only names and descriptions first, activate by loading the full `SKILL.md`, and then read scripts or references only as needed. Client guidance also recommends scanning cross-client paths such as `.agents/skills/`, activating skills model-first or via a dedicated tool, preserving active skill content during context compaction, and gating project-level skills behind trust checks when a repository may be untrusted. citeturn7view1turn7view2turn9view3turn31view0turn31view4turn8view9

The most rigorous recommendation is therefore for UAIX.org to publish a **reference UAIX Agent Skill bundle** that teaches skills-compatible agents how to: discover the current UAIX surface from the live catalog and discovery routes; resolve the correct UAI-1 profile, schema, registry entry, example, and field registry; create a candidate UAI-1 packet; validate it via the public validator route; package the result into an Adoption Kit–style first-proof packet or a Conformance Pack–style release packet; and then promote only reviewed facts into **AI Memory** and **Project Handoff**. That recommendation is not speculative hand-waving: UAIX’s own roadmap already names **“SKILL.md folders”** as part of the capability-package sidecar space it wants to evaluate as **optional UAI evidence mappings**, while explicitly stating that such work is **not** current package-profile support, authorization, adapter support, SDK support, CLI support, certification, endorsement, or a conformance claim. citeturn23view0turn23view1turn33view0turn29view4turn30view2turn30view3

The integration should therefore be marketed with a short, accurate sentence:

> **Agent Skills package UAIX procedures; UAI-1 records UAIX evidence.**

That sentence matches the authoritative boundaries on both sites and avoids the two main failure modes: first, confusing UAIX with a runtime; second, confusing a passing validator result with certification or blanket ecosystem support. UAIX repeatedly states that a validator result is evidence for the reviewed packet and named scope, not a certification surface or broad support claim. citeturn22view7turn17view6turn24view0turn30view4

## Source basis and analytical fit

The authoritative UAIX pages relevant to this integration are the **UAI-1 specification**, **Schemas**, **Registry**, **Examples**, **Validator**, **API Reference**, **Adoption Kit**, **Conformance Pack**, **AI Memory**, **Project Handoff**, **Implementations**, **Roadmap**, and the **Agentic Harness** guide. The API Reference is especially important, because UAIX says implementers should start from the live catalog and discovery routes and use the live OpenAPI export and starter requests instead of reverse-engineering the site or relying on stale notes. citeturn5view0turn23view0turn23view1

From the UAI-1 specification and supporting records, the current published UAI-1 profile family is:

| Profile | Role |
|---|---|
| `uai.intent.request.v1` | Explicit request against a declared subject. citeturn35view1turn34view3 |
| `uai.intent.response.v1` | Direct results, acknowledgements, and accepted async handoffs. citeturn35view1turn34view3 |
| `uai.capability.statement.v1` | Publicly reviewable capability declarations. citeturn35view1turn34view3 |
| `uai.error.v1` | Typed, machine-readable failure records. citeturn35view1turn34view3 |
| `uai.conformance.result.v1` | Exported validator evidence. citeturn35view1turn34view3 |
| `uai.task.status.v1` | Visible async progress and completion state. citeturn35view1turn34view3 |

UAIX also makes the envelope model explicit. The field registry says the stable root order is `uai_version`, `profile`, `message_id`, `source`, `target`, `conversation`, `delivery`, `trust`, `body`, `provenance`, `integrity`, and `extensions`; the schema page explains that the envelope keeps identity, workflow state, trust posture, auditability, provenance, integrity, and extension metadata explicit across profiles. citeturn10view0turn35view5

The public machine surface is unusually mature for a standards site. The API Reference enumerates discovery GET routes for catalog, discovery, schemas, registry, field registry, examples, transport bindings, trust channels, error registry, conformance levels, roadmap, OpenAPI, adoption kit, and conformance pack; it also identifies execution POST routes for validation and mock exchange. The same page gives starter call patterns in curl, PowerShell, Python, and TypeScript, including a Python flow that fetches the adoption kit, catalog, example packet, validator result, and mock exchange response without depending on a published SDK. citeturn23view0turn23view1turn23view2turn23view3

UAIX’s implementation and roadmap material strongly constrain how the integration should be worded. The current public implementation scope is intentionally narrow: the published implementation tracks are **WordPress Publication Track** and **.NET Bridge Track**. UAIX treats only a named implementation-track record plus release-trail entry as a public support claim; validated packets and release-ready packets are necessary evidence, but not the same thing as public support. The roadmap further says UAIX is **not** an agent runtime, scheduler, tool bus, hosted memory platform, official adapter set, SDK, CLI, or certification program today. citeturn4view10turn6view18turn24view0turn30view1turn30view6

Agent Skills fit cleanly into that boundary because they are a procedure format, not a runtime claim. The skills specification requires YAML frontmatter and Markdown body in `SKILL.md`; `name` and `description` are required, `license`, `compatibility`, `metadata`, and `allowed-tools` are optional; and the description should say both what the skill does and when to use it. The `description` field is capped at **1024 characters**, the `name` at **64 characters**, and the skill directory can optionally contain `scripts/`, `references/`, and `assets/`. citeturn9view3turn9view4turn8view1turn8view2

The conceptual responsibility split is therefore straightforward:

| Layer | Best responsibility |
|---|---|
| **Agent Skills** | Discovery metadata, activation behavior, task-specific instructions, reusable scripts, references, assets, and eval-driven refinement. citeturn7view1turn7view6turn7view8turn9view7 |
| **UAIX / UAI-1** | Public exchange contract, identity/trust/provenance envelope, profile registry, validator-backed evidence, first-proof onboarding packet, reusable release packet, AI Memory, and Project Handoff. citeturn5view0turn10view0turn34view3turn22view7turn33view0turn29view4turn19view2turn20view5 |
| **Agent runtime / harness** | Planning, orchestration, tool calls, retries, approvals, runtime memory, traces, and live execution control. UAIX’s harness guide says those concerns belong in the harness, while UAI keeps the reviewed portable evidence. citeturn21view0turn21view3 |

That split is not merely elegant. It is directly aligned with the most conservative reading of both standards, which matters because the roadmap explicitly warns against widening support language before the relevant public evidence exists. citeturn16view14turn30view6

## Data flow and reference architecture

The concrete integration should follow a **discover → activate → resolve → validate → package → hand off** pattern. The first three steps come from Agent Skills’ discovery and progressive-disclosure model; the latter three come from UAIX’s proof path, validator, Adoption Kit, Conformance Pack, AI Memory, and Project Handoff guidance. Agent Skills says discover names and descriptions first, activate the full skill only on demand, and load bundled resources only as needed. UAIX says the proof path is profile → schema/example/registry → validator result → implementation or handoff record, and both the Adoption Kit and the Conformance Pack repeat that same order. citeturn7view1turn7view2turn7view6turn5view0turn5view0turn23view1turn33view0turn29view4

The reference sequence below is the cleanest end-to-end architecture for a UAIX Agent Skill.

```mermaid
sequenceDiagram
    autonumber
    participant User
    participant Runtime as Skills-compatible runtime
    participant Skill as UAIX Agent Skill
    participant UAIX as UAIX public routes
    participant Memory as AI Memory / Project Handoff

    User->>Runtime: Ask for UAIX/UAI-1 work
    Runtime->>Runtime: Discover .agents/skills and load name+description
    Runtime->>Skill: Activate SKILL.md
    Skill->>UAIX: GET /wp-json/uaix/v1/catalog
    Skill->>UAIX: GET /wp-json/uaix/v1/discovery
    Skill->>UAIX: GET schema/registry/example for selected profile
    Skill->>Runtime: Assemble candidate UAI-1 message
    Skill->>UAIX: POST /wp-json/uaix/v1/validate
    UAIX-->>Skill: uai.conformance.result.v1 or validation result
    opt first-proof or response inspection
        Skill->>UAIX: GET /wp-json/uaix/v1/adoption-kit
        Skill->>UAIX: POST /wp-json/uaix/v1/mock-exchange
    end
    opt broader release evidence
        Skill->>UAIX: GET /wp-json/uaix/v1/conformance-pack
    end
    Skill->>Memory: Write reviewed facts only
    Memory-->>Runtime: Updated AGENTS.md / readme.human / .uai / AI Memory
```

A second way to see the same flow is as a data-contract ladder:

| Stage | Primary source of truth | Output |
|---|---|---|
| Skill discovery | `.agents/skills/` catalog and `SKILL.md` frontmatter. citeturn7view2turn9view3 | Skill name, description, location, optional compatibility notes. |
| Skill activation | `SKILL.md` body and referenced resources. citeturn7view1turn31view4 | Procedural guidance for UAIX-specific work. |
| Contract resolution | UAIX catalog, discovery, registry, schemas, examples, field registry. citeturn23view1turn23view3turn34view3turn10view0 | Correct profile and exact machine-readable artifacts. |
| Packet creation | UAI-1 envelope plus profile-specific body. citeturn10view0turn26view0turn26view2turn27view0 | Candidate message. |
| Validation | `POST /wp-json/uaix/v1/validate`. citeturn23view3turn22view2 | Validator result or conformance record. |

Why This File Exists

This is a memory-system evidence 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 file is memory-system evidence. It records source history, archive transfer, intake disposition, or another piece of provenance that should be retrievable without becoming an unsupported public claim.

Structure

The file is structured around these visible headings: UAIX.org and Agent Skills Integration Report; Executive summary; Source basis and analytical fit; Data flow and reference architecture; Proposed UAIX Agent Skill bundle; scripts/fetch_catalog.py; Pseudocode: resolve live UAIX machine surface before route-specific work.; scripts/validate_uai.py. 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-117 (primary)
• Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "UAIX Org And Agent Skills Integration Report",
    "source_site":  "uaix.org",
    "source_url":  "https://uaix.org/",
    "canonical_url":  "https://aiwikis.org/uaix/files/raw-system-archives-uaix-agent-file-handoff-retired-source-archive-2026-1850145c/",
    "source_reference":  "raw/system-archives/uaix/agent-file-handoff/retired-source-archive-2026-06-13/2026-05-06/Improvement/agent-skills-integration/UAIX.org and Agent Skills Integration Report 2.md",
    "file_type":  "md",
    "content_category":  "memory-file",
    "content_hash":  "sha256:1850145c5d856814f01ad748f1ccf22f41c42485d625a582c696f5a15b792c9f",
    "last_fetched":  "2026-06-22T01:56:21.9510185Z",
    "last_changed":  "2026-05-06T22:33:37.3508317Z",
    "import_status":  "unchanged",
    "duplicate_group_id":  "sfg-117",
    "duplicate_role":  "primary",
    "related_files":  [

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