Processed Improvement Inventory

This inventory processes the active Improvement reports as local review inputs. It does not publish them, approve their claims, or widen LlmWikis.org support boundaries.

Metadata

Field	Value
Source site	llmwikis.org
Source URL	https://llmwikis.org/
Canonical AIWikis URL	https://aiwikis.org/llmwikis/files/raw-system-archives-llmwikis-improvement-processing-2026-05-01-processed-6ad15c9e/
Source reference	raw/system-archives/llmwikis/improvement-processing/2026-05-01/PROCESSED_IMPROVEMENT_INVENTORY.md
File type	md
Content category	memory-file
Last fetched	2026-06-22T01:56:21.9510185Z
Last changed	2026-05-01T19:46:40.6681768Z
Content hash	sha256:6ad15c9e40ccc4a886c5cadb114c1eff92ced772f14852572fd64d8fbe2c22f2
Import status	unchanged
Raw source layer	data/sources/llmwikis/raw-system-archives-llmwikis-improvement-processing-2026-05-01-processed-improvement-inventory-m-6ad15c9e40cc.md
Normalized source layer	data/normalized/llmwikis/raw-system-archives-llmwikis-improvement-processing-2026-05-01-processed-improvement-inventory-m-6ad15c9e40cc.txt

Current File Content

Structure Preview

• Processed Improvement Inventory
• Repository Structure Found
• Existing Related Content Located
• Source File List And Disposition
• Per-File Summaries
• Governed Integration of LLM Wiki, UAI AI Memory, UAI Project Handoff, and Cursor
• Integrating LLM Wiki, UAI AI Memory, UAI AI Project Handoff, and a Codex Coding Agent
• Layered Publishing Architecture with LLM and Claude Code
• Layered Publishing Architecture with LLM and Gemini
• LLM Wiki, UAI AI, Gemini Integration Report
• LLM Wiki, UAI, Claude Integration Report
• LLM Wiki, UAIX, Codex Integration Report
• LLM Wiki, UAIX, Cursor Integration Report
• Overlap
• Contradictions Or Verification Needs
• Recommended Site Changes
• Recommended Documentation And Handoff Changes
• Risks And Review Requirements

Raw Version

---
status: staged-review
created: 2026-05-01
source_bucket: agent-file-handoff/Improvement
public_truth: false
---

# Processed Improvement Inventory

This inventory processes the active Improvement reports as local review inputs. It does not publish them, approve their claims, or widen LlmWikis.org support boundaries.

## Repository Structure Found

- This is a mixed WordPress Studio repository with WordPress core, a custom public theme, a structured content plugin, root discovery files, local scripts, and repo-local handoff files.
- Public pages are seeded through `wp-content/themes/llmwikis-knowledge-theme/functions.php` and helper content in `wp-content/themes/llmwikis-knowledge-theme/inc/reference-pages.php`.
- Active theme: `wp-content/themes/llmwikis-knowledge-theme/`.
- Structured content plugin: `wp-content/plugins/llmwikis-core/`.
- Discovery files found: `llms.txt`, `robots.txt`, and `sitemap.xml`. No `llms-full.txt` file was found.
- Handoff files already exist: `AGENTS.md`, `readme.human`, and typed `.uai/` records.

## Existing Related Content Located

- What is an LLM Wiki: `/what-is-an-llm-wiki/`
- Metadata standard: `/metadata-standard/`
- Trust model: `/trust-model/`
- For AI agents: `/for-ai-agents/`
- Security and privacy: `/security-and-privacy/`
- Tooling landscape: `/implementations/tooling-landscape/`
- LLM Wiki vs AI Memory: `/llm-wiki-vs-ai-memory/`
- UAI/UAIX integration: `/reports/using-llm-wiki-with-uai/`, `/uaix-bridge/`, `/standards/`
- Project Handoff: `/guides/agent-handoff-patterns/`
- Machine discovery: `llms.txt`, `sitemap.xml`, `robots.txt`
- Agent guidance: `AGENTS.md`, `readme.human`, `.uai/*.uai`

## Source File List And Disposition

The intake helper was run on 2026-05-01 and indexed eight files. The original reports were left in `agent-file-handoff/Improvement/` for human review alongside this package. After acceptance, move the original eight reports to `agent-file-handoff/Archive/` and rerun intake.

| Source file | Disposition |
|---|---|
| `Governed Integration of LLM Wiki, UAI AI Memory, UAI Project Handoff, and Cursor.md` | `apply-now` into staged planning; `convert-into-roadmap-progress` for private Cursor/middleware ideas |
| `Integrating LLM Wiki, UAI AI Memory, UAI AI Project Handoff, and a Codex Coding Agent.md` | `apply-now` into staged Codex guide; vendor-specific claims need fact-checking |
| `Layered Publishing Architecture with LLM and Claude Code.md` | `apply-now` into review-gated publication draft; Claude-specific claims need fact-checking |
| `Layered Publishing Architecture with LLM and Gemini.md` | `apply-now` into layered memory and risks drafts; Gemini-specific claims need fact-checking |
| `LLM Wiki, UAI AI, Gemini Integration Report.md` | `convert-into-roadmap-progress`; not public-ready without source review |
| `LLM Wiki, UAI, Claude Integration Report.md` | `convert-into-roadmap-progress`; not public-ready without source review |
| `LLM Wiki, UAIX, Codex Integration Report.md` | `apply-now` into staged Codex and handoff drafts; block overclaims around automatic memory/write-back |
| `LLM Wiki, UAIX, Cursor Integration Report.md` | `convert-into-roadmap-progress`; block continuous auto-publish interpretations |

## Per-File Summaries

### Governed Integration of LLM Wiki, UAI AI Memory, UAI Project Handoff, and Cursor

- Separates LLM Wiki, AI Memory, Project Handoff, Cursor, and UAI-1 into distinct layers.
- Recommends middleware-centered integration where Cursor loads compact memory first, queries wiki only as needed, stages changes, and promotes after review.
- Correctly notes LlmWikis.org is a handbook/starter-bundle site today, not a public API, public MCP, or open editing service.
- Useful ideas: private MCP gateway, staged promotion jobs, contradiction records, manifest refresh, and UAI-1 validation for evidence-bearing flows.
- Needs fresh Cursor official-doc verification before public use.

### Integrating LLM Wiki, UAI AI Memory, UAI AI Project Handoff, and a Codex Coding Agent

- Separates LLM Wiki as durable memory, AI Memory as compact working context, Project Handoff as transfer context, and Codex as execution.
- Recommends reviewed export from wiki pages into Project AI Memory or Project Handoff before agent execution.
- Provides a useful mapping from wiki content to `.uai` architecture, decisions, progress, constraints, operations, and test-plan records.
- Reinforces that wiki memory is background until reviewed and promoted into named project artifacts.
- Current OpenAI/Codex claims require official-doc verification before publication.

### Layered Publishing Architecture with LLM and Claude Code

- Recommends review-gated publication rather than autonomous live publishing.
- Suggests `AGENTS.md` as the portable front door and a thin `CLAUDE.md` compatibility pointer.
- Provides useful publication checks, risk categories, and metrics.
- Supports source-policy, metadata, trust labels, and human review gates.
- Claude hooks/MCP/Agent SDK claims require verification.

### Layered Publishing Architecture with LLM and Gemini

- Recommends LLM Wiki as durable knowledge, AI Memory as generated export, Project Handoff as transfer packet, and Gemini as execution surface.
- Emphasizes that local `.uai` records are useful operational context, not formal UAI-1 conformance evidence.
- Identifies risks around memory drift, secret leakage, over-trusting agent mode, cost, auth ambiguity, and API churn.
- Suggests phased rollout: foundation, operational memory, coding-agent integration, handoff discipline, evidence/scale.
- Gemini feature, pricing, privacy, beta, and API claims require verification.

### LLM Wiki, UAI AI, Gemini Integration Report

- Explains LLM Wiki as durable maintained markdown rather than query-time retrieval alone.
- Describes raw/wiki/index/log patterns and trust labels.
- Frames Gemini as execution layer with memory and handoff concepts.
- Useful as background for roadmap and adapter planning.
- Contains broad industry and vendor claims that are not public-ready without primary-source review.

### LLM Wiki, UAI, Claude Integration Report

- Explains raw/wiki/site separation, two-step ingest, metadata, trust labels, contradiction edges, index/log navigation, AI Memory, and Project Handoff.
- Supports compact `AGENTS.md` and `readme.human` as a human briefing.
- Suggests Claude compatibility through `CLAUDE.md`, local memory, MCP retrieval, and staged write-back.
- Warns against context pollution and stale local agent memory.
- Includes creator attribution and vendor claims that should not be copied into public copy without review.

### LLM Wiki, UAIX, Codex Integration Report

- Covers durable knowledge, compiled wiki pages, deterministic routing, contradiction handling, and Codex execution.
- Reinforces immutable `raw/`, compiled `wiki/`, `index.md`, `log.md`, metadata, status labels, and staged review.
- Useful for a Codex guide focused on handoff front doors, source citation, uncertainty, targeted checks, and no direct publication.
- Codex internals, models, modes, privacy, and performance claims need official OpenAI verification.
- UAI-1 should be framed as evidence for public exchange/support claims, not a mandatory wrapper for every internal wiki operation.

### LLM Wiki, UAIX, Cursor Integration Report

- Emphasizes context amnesia, AI Memory startup packets, `AGENTS.md`, raw/wiki/schema separation, `llms.txt`, Cursor execution, and handoff acceptance.
- Useful for handoff rules and a future coding-agent adapter roadmap.
- Includes speculative automation ideas such as background checkpoints, transcript ingestion, memory hooks, and continuous write-back.
- Conflicts with current support boundaries unless framed as private, gated, future work.
- Vendor and creator-attribution claims need review before public use.

## Overlap

- All reports support the same core separation: LLM Wiki is durable reviewed knowledge; AI Memory is compact operating context; Project Handoff is transfer context; coding agents execute tasks.
- All reports support staged promotion, source citations, review gates, trust labels, and contradiction preservation.
- All reports support repo-local front doors such as `AGENTS.md`, `readme.human`, and `.uai` records.
- The best shared recommendation is governed drafts, PRs, changelogs, and staged content rather than direct production publishing.

## Contradictions Or Verification Needs

- Some reports imply public MCP, automatic write-back, hooks, background memory, or live adapters. Current LlmWikis constraints do not support those claims publicly.
- Some reports imply `.uai` files are formal UAI-1 conformance evidence. Current repo rules treat them as operational context unless promoted into formal evidence.
- Vendor-specific product capabilities for Codex, Cursor, Claude, and Gemini need fresh official-source verification.
- Public copy must avoid creator-name attribution for the LLM Wiki concept.
- Numerical confidence scoring and automated decay should remain roadmap/background until qualitative labels, contradiction records, and human review queues are mature.
- Do not add `llms-full.txt` without a separate discovery-process decision.

## Recommended Site Changes

- Stage a new guide candidate for `/guides/using-llm-wiki-with-codex/`.
- Stage a guide or expansion explaining LLM Wiki, AI Memory, and Project Handoff together.
- Stage a governance page or section on review-gated publication.
- Add concise section candidates for `/for-ai-agents/`, `/security-and-privacy/`, and `/guides/agent-handoff-patterns/`.
- Keep adapter details for Cursor, Claude, Gemini, and Codex in staged/internal form until fact-checked.

## Recommended Documentation And Handoff Changes

- Keep existing `AGENTS.md`, `readme.human`, and `.uai` files; they already satisfy the baseline requested files.
- Propose minimal future `.uai/progress.uai` and `.uai/strategy.uai` notes after human acceptance.
- Keep tool-specific files thin and redirect back to `AGENTS.md`.
- Do not add watchers, daemons, scheduled jobs, CI pickup, public MCP, public write APIs, automatic publication, certification, SDK, CLI, or official adapter claims.

## Risks And Review Requirements

- Public overclaim risk: no public MCP, public write APIs, automatic publishing, certification, official adapters, or automatic sync without evidence.
- Authority risk: UAIX remains canonical for UAI-1, AI Memory, Project Handoff, schemas, validator behavior, roadmap, and governance.
- Security risk: no secrets, credentials, private keys, private transcripts, raw regulated/customer data, or sensitive legal material.
- Discovery risk: staged drafts and private handoff files must not enter `llms.txt` or `sitemap.xml` until accepted.
- Source risk: vendor/tool claims need primary-source verification before public use.

Why This File Exists

This is a memory-system evidence file from llmwikis.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: Processed Improvement Inventory; Repository Structure Found; Existing Related Content Located; Source File List And Disposition; Per-File Summaries; Governed Integration of LLM Wiki, UAI AI Memory, UAI Project Handoff, and Cursor; Integrating LLM Wiki, UAI AI Memory, UAI AI Project Handoff, and a Codex Coding Agent; Layered Publishing Architecture with LLM and Claude Code. 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-511 (primary)
• Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "Processed Improvement Inventory",
    "source_site":  "llmwikis.org",
    "source_url":  "https://llmwikis.org/",
    "canonical_url":  "https://aiwikis.org/llmwikis/files/raw-system-archives-llmwikis-improvement-processing-2026-05-01-processed-6ad15c9e/",
    "source_reference":  "raw/system-archives/llmwikis/improvement-processing/2026-05-01/PROCESSED_IMPROVEMENT_INVENTORY.md",
    "file_type":  "md",
    "content_category":  "memory-file",
    "content_hash":  "sha256:6ad15c9e40ccc4a886c5cadb114c1eff92ced772f14852572fd64d8fbe2c22f2",
    "last_fetched":  "2026-06-22T01:56:21.9510185Z",
    "last_changed":  "2026-05-01T19:46:40.6681768Z",
    "import_status":  "unchanged",
    "duplicate_group_id":  "sfg-511",
    "duplicate_role":  "primary",
    "related_files":  [

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