Skip to content
aiWikis.org

LlmWikis Progress

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

Metadata

FieldValue
Source siteaiwikis.org
Source URLhttps://aiwikis.org/
Canonical AIWikis URLhttps://aiwikis.org/files/aiwikis/raw-system-archives-llmwikis-internal-memory-reorg-2026-05-01-uai-progre-e3fd6033/
Source referenceraw/system-archives/llmwikis/internal-memory-reorg/2026-05-01/.uai/progress.uai
File typeuai
Content categoryuai-system
Last fetched2026-05-02T01:47:31.8867765Z
Last changed2026-04-29T13:35:18.0182259Z
Content hashsha256:e3fd6033b818bc4066a61f9815a5ee6052202e75074d20ac60013d12a395f2c3
Import statusunchanged
Raw source layerdata/sources/aiwikis/raw-system-archives-llmwikis-internal-memory-reorg-2026-05-01-uai-progress-uai-e3fd6033b818.uai
Normalized source layerdata/normalized/aiwikis/raw-system-archives-llmwikis-internal-memory-reorg-2026-05-01-uai-progress-uai-e3fd6033b818.txt

Current File Content

Structure Preview

  • LlmWikis Progress
  • Recently Completed
  • Current Focus
  • Known Blockers
  • Next Work
  • Done Means For This Production Pass

Raw Version

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: progress
title: "LlmWikis Progress"
created: "2026-04-26"
updated: "2026-04-28"
author: "LlmWikis maintainers"
version: 32
---

# LlmWikis Progress

## Recently Completed

- Added a shared handbook page layout pass after static render review showed ordinary pages still wasting width and table content clipping on mobile. Seeded handbook tables now use the full desktop reading surface with fixed columns and compress into wrapped, phone-width tables on small screens, while page cards and notices keep the dense reference-site rhythm established on the homepage.
- Added a second LlmWikis homepage layout pass for below-the-fold rhythm: build-step tiles now use a balanced six-column wide layout with tablet fallbacks, implementation modules settle into an even four-by-two grid, status/evidence/discovery cards are tighter, and the reference-site posture split keeps two clear columns on desktop while preserving mobile reflow.
- Extended the LlmWikis homepage screen-real-estate pass after visual review showed the homepage still reading as a narrow center column. The active theme now reduces desktop gutters, widens the homepage shell, rebalances the hero and search band, turns the four Handbook routes into a true wide desktop grid, tightens route-card icon and list spacing, preserves mobile reflow, and bumps the theme asset version for cache refresh.
- Reworked the LlmWikis front-end shell after visual review showed the reference layout still wasting viewport width and the old contrast control no longer behaving predictably. The active theme now has a final liquid-width cascade, denser handbook grids and section spacing, a persistent Light/Dark toggle with labels that reflect the active theme, and a deeper dark palette that inverts the civic/reference surfaces without returning to the obsolete black portal design.
- Stabilized the LlmWikis public layout after upload review showed dark legacy styling and brittle homepage grids leaking through the newer civic/reference design. The active theme now has a final layout cascade that restores light page surfaces, resilient homepage grids, readable hero/search/status/topic/process/agent sections, safer mobile header/menu behavior, scroll-safe media and tables, and a bumped asset version for cache refresh.
- Drained the current already-dispositioned LlmWikis archive files into AIWikis.org `raw/system-archives/llmwikis/` as long-term system memory after explicit human direction. LlmWikis keeps normal active intake and disposition behavior, and AIWikis archive copies do not become canonical LlmWikis or UAIX authority.
- Applied the updated LlmWikis brand image drops from `agent-file-handoff/Improvement/` at human request: regenerated theme/site favicon assets, app icons, thumbnails, and 1200x630 social-share imagery from `LLM Wikis Logo.png` and `LLM Wikis.png`; replaced the old SVG logo references; archived the processed source files under `agent-file-handoff/Archive/2026-04-28/Improvement/`; and refreshed active intake to 0 files.
- Updated LlmWikis public Agent Handoff guidance with the AIWikis.org dogfood finding: starter bundles and WordPress handoff exports should be checked for actual `agent-file-handoff/` folders, the loaded `.uai` handoff set, source manifests, and release evidence before claiming Project Handoff dogfood. LlmWikis continues to consume the UAIX-published `uai1-project-handoff.zip` rather than treating a local copy as canonical.
- Clarified the package ownership boundary from the other side: LlmWikis.org is the source publisher for `llm-wiki-starter-bundle.zip`, while AIWikis.org and other dogfood sites should consume that source-published bundle instead of creating site-branded starter clones.
- Corrected the LlmWikis launch posture for first public release: updated current handoff state from live-site language to prelaunch/launch-ready language, removed the obsolete nested What Is 301 redirect, retired `/start-here/what-is-an-llm-wiki/` from managed pages as prelaunch route debt, kept `/what-is-an-llm-wiki/` canonical, hardened production metadata to `https://llmwikis.org`, added baseline security headers, and fixed the dynamic starter ZIP route so it returns a true 200 response with ZIP/noindex/nosniff headers.
- Added a deep LLMWikis.org public-reference pass: first-class pages for `/what-is-an-llm-wiki/`, `/why-llm-wikis/`, `/how-to-build-an-llm-wiki/`, `/llm-wiki-structure/`, `/starter-template/`, `/content-types/`, `/trust-model/`, `/metadata-standard/`, `/for-ai-agents/`, `/llm-wiki-vs-rag/`, `/llm-wiki-vs-ai-memory/`, `/checklist/`, `/maturity-model/`, and `/security-and-privacy/`; refreshed homepage, Explore, Guides, header, footer, examples, governance, security, `llms.txt`, and `sitemap.xml`; moved the public managed route count to 74; and retired the old nested What Is route before launch instead of preserving it as public SEO debt.
- Added a canonical LLM Wiki starter-template and bundle architecture under `wp-content/themes/llmwikis-knowledge-theme/inc/`: template registry metadata, starter structure, bundle manifest, dynamic ZIP generation, validation for duplicate IDs/paths, unsafe paths, empty bundles, missing template references, and tests that fail if visible examples drift from downloaded files.
- Added targeted regression scripts `scripts/test-llmwikis-starter-bundle.php` and `scripts/Test-LlmWikisPublicSurface.ps1`; the public-surface PowerShell check passes locally. Refreshed `[local path redacted]` packages and checksums after the broad public-surface change. Direct PHP syntax/test execution is blocked because `php` is not on PATH, and Studio runtime checks remain blocked by the local Studio `EPERM` migration issue.
- Processed `agent-file-handoff/Improvement/Improving the LlmWikis Contrast Theme.md` as a site-improvement audit. Disposition: convert into roadmap/progress state for a future focused front-end pass covering tokenized color and typography, stronger focus states, header/source-map simplification, prose measure, responsive tables/code, footer density, and WCAG contrast/reflow checks; do not implement or publish the raw report as public copy in this test-policy pass.
- Dispositioned the active duplicate `agent-file-handoff/Content/Llm Wiki Navigation Strategy Guide.md` intake file as already processed on 2026-04-28, with no new public content work required, and moved the duplicate copy into `agent-file-handoff/Archive/2026-04-28/Content/`.
- Added `.uai/test-plan.uai` as the LlmWikis handoff file for test-selection policy. Ordinary prompt work now defaults to targeted checks tied to changed files, routes, discovery files, or handoff records, while full package/runtime sweeps and ZIP refreshes are reserved for release/package work, broad public-surface changes, or explicit full-check requests. Full builds now also include source-document hygiene for `AGENTS.md`, `readme.human`, `.uai` files, progress/strategy/public-surface records, and active Markdown source records while keeping private source files out of discovery and upload ZIPs.
- Processed the 2026-04-28 active Content intake set `Developing LLM Wiki Operations Content.md`, `LLM Wiki Architecture and Examples.md`, and `Llm Wiki Navigation Strategy Guide.md` into public handbook improvements rather than raw report pages. Applied deeper specification-style content across the architecture, operations, navigation, three-layer architecture, RAG vs compilation, ingest, query, lint, index/log, AGENTS.md/schema, personal/team examples, memory lifecycle, two-step ingest, graph navigation, page schema, tooling landscape, Source Policy, and Page Templates routes. Kept MCP/public automation, hosted ingestion, numeric confidence, open editing, and enterprise-service claims gated as planned or implementation-specific work. Archived the processed source files, refreshed active intake to 0 files, and rebuilt the local LlmWikis publish ZIPs and checksums.
- Processed the 2026-04-28 active improvement report `Deep Research Report on llmwikis.org Content Refinement.md` as site-improvement input, not public report copy. Applied it into deeper source-aware public content across UAI-1 explainers, message-model guidance, validator errors, validation workflow, Validator Workbench, AGENTS.md and handoff pages, personal/team examples, UAI-1 comparisons, model/evaluation scaffolding, the UAIX bridge, and the Using LLM Wiki with UAI report; added quality/source-status metadata to shared page status blocks; archived the source file under `agent-file-handoff/Archive/2026-04-28/Improvement/`; and refreshed intake to 0 active files.
- Processed the 2026-04-28 active intake set from `agent-file-handoff/Content/` and `agent-file-handoff/Improvement/`. Dispositions: `Creating LLM Wiki Source Policy.md`, `Creating LLMWikis Start Here Page.md`, `LLM Wiki Architecture.md`, and `Revised Publication Report for What Is an LLM Wiki.md` were applied into public handbook and policy pages after tightening source/support claims; `LLM Wiki Content Refinement Report.md` was converted into page improvements plus roadmap/progress state, not public report copy. The public `/content-quality/` route family is now hard-retired with a permanent `410 Gone` handler, header/footer links were removed, and the old audit page functions were removed from the active theme.
- Added `agent-file-handoff/Archive/` as the dead archive for already-dispositioned intake files. Routine AI intake must ignore archived files unless a human explicitly names one or moves it back into `Content/` or `Improvement/`.
- Removed the recommendation for bucket-local README files from the Agent File Handoff pattern; durable AI instructions now stay in `AGENTS.md` and `.uai` files, while human-facing explanation stays in root `readme.human`.
- Archived the already-processed LlmWikis intake files under `agent-file-handoff/Archive/2026-04-27/Content/` and `agent-file-handoff/Archive/2026-04-27/Improvement/` after their useful information was converted into public page, roadmap, progress, and strategy state.
- Processed current intake files from `agent-file-handoff/Content/` and `agent-file-handoff/Improvement/`. Dispositions: `Schema Engineering for Durable LLM Wikis.md` is already public at `/schema-engineering/durable-llm-wikis/` and was used to add a compact schema-measurement section; `LLM Wiki Content Refinement Report.md` was converted into roadmap/progress state for schema gallery, epistemic governance, agent-history adapters, graph/API exposure, and confidence-label work; `UAIX SEO Improvement Report.md` is UAIX-side improvement input and should not become LlmWikis public content.
- Expanded the public Roadmap page with explicit gates for schema-gallery templates, qualitative confidence/review labels, agent-history adapters, and graph/API exposure while keeping public MCP, automated ingestion, and broad AI search gated until implementation, security review, and human policy exist.
- Added a larger UI/UX and SEO handbook expansion: five source-bounded technical routes for `/architecture/memory-lifecycle/`, `/operations/two-step-ingest-pipeline/`, `/navigation/graph-navigation/`, `/schema-engineering/page-schema-standard/`, and `/implementations/tooling-landscape/`; refreshed the homepage route modules; added section-aware structured data for managed pages; removed retired `/content-quality/` audit routes from `llms.txt` and `sitemap.xml`; and bumped `llmwikis-knowledge-theme` to reseed managed pages.
- Aligned reader-facing Agent Handoff copy in the active theme with the current `agent-file-handoff/Content/` and `agent-file-handoff/Improvement/` bucketed intake pattern, replacing older single-folder intake wording while preserving the no-watcher/no-daemon boundary.
- Split Agent File Handoff into `agent-file-handoff/Content/` for candidate public/editorial material and `agent-file-handoff/Improvement/` for audits, SEO reports, QA findings, bug notes, roadmap suggestions, and other site-improvement work. The legacy `intake/` folder is compatibility-only.
- Demoted the content-quality audit from public handbook/discovery material into internal improvement/backlog state; the audit is work to do on the site, not public site content by default.
- Rebuilt the LlmWikis homepage and theme presentation around an informational non-profit/reference-site model: liquid layout, compact navigation/search/status, dense route indexes, practical first-working sequence, source-boundary tables, and a new paper/ink/sage/rust/blue civic palette with no Bootstrap dependency.
- Added a Screen Real Estate density pass to the LlmWikis public theme: compacted documentation headers, status panels, cards, metrics, reading grids, and footer spacing; replaced the sparse Schema Engineering hub with a dense operating-contract page carrying control areas, artifact matrix, starter schema outline, failure modes, and next-page routing.
- Added root `readme.human` as the human-facing Project Handoff companion to `AGENTS.md`, aligned public LlmWikis handoff guidance with UAIX, and kept the file out of public discovery outputs.
- Tightened Agent File Handoff from passive indexing into immediate review/disposition: non-executable drops are now `needs-agent-review`, and agents must inspect, summarize, and disposition each listed file before unrelated broad work.
- Processed the legacy intake draft `Schema Engineering for Durable LLM Wikis.md` into the public `/schema-engineering/durable-llm-wikis/` route, with managed-page seeding, per-route meta description, `llms.txt`, and `sitemap.xml` coverage.
- Historical intermediate: simplified Agent File Handoff to a single `agent-file-handoff/intake/` drop folder plus `.uai/intake-index.uai`, removing queue-folder, manifest, watcher, and daemon guidance from the base pattern. Current active intake now uses `Content/`, `Improvement/`, and ignored `Archive/`.
- Historical intermediate: clarified active local Agent File Handoff intake as AGENTS.md-triggered chat-start behavior without out-of-chat automation. Current dogfood checks active buckets during normal handoff load.
- Historical intermediate: added active local Agent File Handoff intake with `scripts/Invoke-UaiFileIntake.ps1`, `.uai/file-handoff.uai`, and `.uai/intake-index.uai`, aligned to UAIX's canonical Agent File Handoff guidance. Current dogfood uses bucketed intake and the dead archive.
- Migrated the retired `docs` Markdown planning set: public reader value now lives in handbook routes and `/reports/using-llm-wiki-with-uai/`; internal strategy, content-quality, platform, roadmap, and implementation value now lives in `.uai/strategy.uai`, handoff files, improvement/backlog state, and roadmap state. Raw source Markdown docs were removed from the repository and packaging.
- Completed a site-wide content and SEO readiness sweep: replaced placeholder-style model, benchmark, community, governance, A2A comparison, model-comparison, benchmark-contamination, and OWASP security copy with useful source-bounded guidance; expanded per-route meta descriptions; added canonical/social/JSON-LD metadata; refreshed `llms.txt` and a 60-route `sitemap.xml`; and bumped `llmwikis-knowledge-theme` for managed page reseeding.
- Initially implemented the April 2026 content-quality audit as public routes, then corrected that interpretation: those audit outputs are internal improvement/backlog material unless a specific piece is deliberately converted into public content.
- Added `/reports/` and published `/reports/using-llm-wiki-with-uai/` as a non-normative HTML report article, with UAIX.org kept canonical for UAI-1 and Project Handoff.
- Updated header, footer, homepage cards, root `llms.txt`, and `sitemap.xml` for Reports; content-quality audit routes were removed from public discovery/navigation and kept as internal improvement state.
- Reframed the launch-ready public site into a practical LLM Wiki creation handbook based on migrated architecture and practical-reference source material.
- Added Start Here, Architecture, Operations, Navigation, Schema Engineering, Examples, and Protocols and Case Studies seeded routes.
- Added public tutorials for What Is an LLM Wiki, the three-layer architecture, RAG vs incremental compilation, ingest, query, lint, index.md/log.md, AGENTS.md/CLAUDE.md design, personal LLM Wiki, and team LLM Wiki.
- Updated the homepage, primary navigation, footer, Explore page, Guides hub, Page Templates, Roadmap, and Related Links to make LLM Wiki creation the public center of gravity.
- Added root `llms.txt`, `robots.txt`, and `sitemap.xml` discovery assets.
- Updated the package script so the source-only handoff archive includes public discovery files.
- Preserved handbook implementation notes in `.uai/progress.uai`, `.uai/stack.uai`, `.uai/architecture.uai`, and package metadata.
- Refreshed WordPress upload ZIP packages and source/site archives under `[local path redacted]`.
- Loaded the UAIX Project Handoff pattern from the UAIX workspace.
- Reframed the public copy, handoff bundle, and publish metadata around LlmWikis as a production package instead of a throwaway preview.
- Aligned LlmWikis package publishing to keep local ZIP outputs in `[local path redacted]` while copying the mirrored UAIX Project Handoff package from `[local path redacted]`.
- Split LlmWikis publish output so the root publish folder contains only the WordPress upload packages, while source-only and full-site archives now live under `[local path redacted]`.
- Clarified that `llmwikis-knowledge-theme` is the only maintained LlmWikis release theme.
- Migrated strategy document value into `.uai/strategy.uai`.
- Created a project-specific `AGENTS.md` plus `.uai/` bundle.
- Created the first LlmWikis production theme baseline.
- Created a structured content plugin for model, benchmark, and protocol records.
- Installed the UAIX Project Handoff prototype plugin for internal package workflows.
- Created an internal UAI-1 handoff export at `wp-content/uploads/uai1-handoffs/exports/llmwikis-org-ai-project-handoff-20260426-195528.uai1.zip`; its bundled validation report is valid with no errors or warnings.
- Started the local WordPress Studio site and restored the missing generated `wp-config.php`.
- Converted project naming to `LlmWikis`, `LlmWikis.org`, and `https://llmwikis.org/` while keeping lower-case `llmwikis-*` handles for code, theme, plugin, CSS, and package paths.
- Removed the copied `.git` binding from `[local path redacted]` so source control can be re-added under the corrected name.
- Turned the UI/UX product blueprint into the current WordPress production surface: ecosystem bar, searchable command-center homepage, task paths, featured knowledge paths, authority boundary, featured reading, and contribution prompt.
- Expanded seeded routes for Explore, Standards, Guides, Comparisons, Glossary, Implementations, UAI-1 explainers, UAI-1 comparisons, agent handoff, validation workflow, concept pages, and Validator Workbench.
- Expanded `llmwikis-core` with structured content types for guides, comparisons, glossary entries, and implementation tools, plus page-type, authority-label, evidence-status, source, and review metadata.
- Added roadmap/doc updates that separate work shipped in WordPress from deferred infrastructure, legal, search, benchmark, contributor, AI, and monetization work.
- Added search/archive/single/404 block templates, a shrinking-header/search-focus script, security/source-policy/template/community pages, model and benchmark template routes, and a richer public roadmap.
- Added `/related-links/` as a footer-linked outgoing source directory for UAIX, standards, protocol, model-provider, benchmark, security, and research references, with explicit non-endorsement and non-normative language.
- Hardened the LlmWikis public route map by moving future structured record rewrites under `/records/...`, forcing a theme seed version bump for missing public child pages, expanding the six formerly broken route targets with meaningful starter content, and polishing homepage, navigation, page status, related-page, and 404 presentation.
- Replaced the prior outline-like homepage with a custom modern documentation portal: dark technical hero, CSS knowledge-graph panel, product-style trust strip, action cards, feature cards, polished metric band, featured reading grid, documentation footer, and stronger internal metadata panels.
- Published WordPress upload ZIP packages to `[local path redacted]` with `checksums.sha256.txt`: `llmwikis-knowledge-theme.zip`, `llmwikis-core.zip`, and `uai1-project-handoff.zip`.
- Published source-only and site-archive bundles under `[local path redacted]`: `llmwikis-handoff-bundle.zip` and `llmwikis-wordpress-site.zip`.

## Current Focus

- Keep informational pages dense and content-forward: first viewports should carry useful reference material, route choices, matrices, examples, status, and source context instead of decorative whitespace or oversized brochure-style panels.
- Keep the final civic/reference CSS cascade authoritative over older dark portal rules until the stylesheet is fully consolidated; maintain liquid page gutters, dense route grids, and the Light/Dark theme contract, and do not reintroduce black-background handbook pages, brittle fixed grids, or mobile-only width clamps that make text wrap one word per line.
- Keep the local site and launch-ready public surface aligned around the LLM Wiki handbook pivot.
- Keep public LLM Wiki creation copy detailed, practical, and source-aware.
- Keep the new technical modules prominent as high-intent SEO landing pages for LLM Wiki ingest pipelines, page schemas, memory lifecycle, graph navigation, and tooling selection.
- Keep schema metrics, qualitative confidence labels, contradiction records, and graph/API exposure framed as reviewable roadmap work unless the implementation and policy are actually live.
- Keep protocol and standards material under the non-normative case-study track.
- Keep UAIX canonical authority separate from LlmWikis explainers.
- Seed enough taxonomy, authority labeling, and policy language for a useful first visitor experience.
- Keep future work marked as planned until it exists.
- Keep Related Links useful as a source map without treating outgoing links as LlmWikis endorsement, freshness guarantees, or UAIX authority.
- Keep content-quality audit work in internal improvement/backlog state until a specific reader-facing page is deliberately approved and source-reviewed.
- Keep Reports clearly non-normative and source-backed, especially when they discuss UAIX/UAI concepts.
- Keep chat-start file intake running before broad work so newly supplied active files are visible and reviewed immediately without relying on private chat memory, watcher/daemon automation, queue-folder workflows, manifests, out-of-chat background processes, or bucket-local README instructions. Keep `Content/` and `Improvement/` distinct so work-to-publish and work-to-fix are not confused, and keep `Archive/` ignored after disposition.
- Keep the source-site archive boundary clear: processed archive files can be pulled into AIWikis long-term memory only by explicit request, and AIWikis copies are not LlmWikis or UAIX authority.
- Keep root `readme.human` aligned with the AI-facing handoff bundle whenever people need new steering guidance.
- Keep targeted verification as the default for ordinary work, and keep full package/runtime sweeps tied to ZIP builds, release gates, broad public-surface changes, or explicit human requests.
- Keep full build/release work responsible for cleaning and updating internal `.md`, `.uai`, and `.human` source files before package output is treated as current.
- Keep clean public page routes separate from future structured record URL spaces so page hubs and child explainers do not conflict with custom post type rewrites.
- Keep the homepage visually distinct from default WordPress article layouts by using the public-interest reference layout, liquid page sections, compact route indexes, and the civic/reference palette while avoiding Bootstrap, jQuery, fake live data, unreviewed support claims, and creator-name attribution for the LLM Wiki concept.

## Known Blockers

- LlmWikis.org is still prelaunch from this workspace's perspective; detailed production hosting, CDN, DNS, and monitoring configuration is not modeled directly in this local workspace.
- A previous typo-named workspace folder remains because its IDE cache files are locked by another process. The active copied workspace is `[local path redacted]`; remove the old folder after the IDE releases the cache.
- Studio CLI currently has the typo-named `[local path redacted]` site registered, but the active corrected `[local path redacted]` directory is not added to Studio yet; `studio wp` runtime checks are blocked until the corrected path is registered or imported.
- Analytics, Search Console, uptime monitoring, legal pages, license choice, privacy terms, and contributor intake need human decisions.
- No live external benchmark APIs or automated ingestion pipelines are connected yet.
- Public multilingual support is not implemented.
- External uptime, backup, CDN, and edge-header monitoring still need operator-owned production review outside this repo.
- Citation-backed conversational search and AI-assisted contribution workflows are not implemented.

## Next Work

- Add real source-backed model and benchmark records.
- Deepen LLM Wiki handbook coverage with starter bundle variants, worked codebase/support-doc/research wiki examples, CI-ready lint recipes, and source-backed tool comparisons.
- Run a focused contrast/accessibility front-end pass from the processed `Improving the LlmWikis Contrast Theme.md` audit: tokenized palette and typography, clearer header/source-map layers, stronger keyboard focus, constrained prose measure, resilient code/table overflow, and mobile reflow checks.
- Keep regenerated favicon, thumbnail, and social-share assets aligned when the LlmWikis source logo files change again.
- Add a source-reviewed schema gallery for codebase, research, support, legal, and team LLM Wiki variants.
- Add private-safe adapter guidance for converting agent session histories into source-reviewed wiki inputs without exposing private transcripts.
- Keep already-dispositioned intake files in `agent-file-handoff/Archive/` so old source material remains human-referenceable without becoming fresh AI work every session.
- Add qualitative confidence, contradiction, supersession, and stale-claim guidance before considering numeric confidence scoring or automated decay models.
- Add public legal/privacy/license/DMCA pages before contributor intake.
- Register the corrected `[local path redacted]` path in Studio or import it as a Studio site so runtime lint, activation, seeded-page, and HTTP smoke checks can run.
- Use `.uai/test-plan.uai` to choose targeted checks for ordinary edits and reserve `scripts/publish-llmwikis-packages.ps1` for ZIP/release work or explicit full-check requests.
- Add grouped search results and citation-backed search only after a source index exists.
- Add page templates in the editor for concept, protocol, comparison, model, benchmark, and implementation records.
- Add a model compare tool and benchmark tables after source-backed data is ready.
- Decide legal/license/privacy/DMCA wording before inviting contributors.
- Add analytics, uptime monitoring, backups, and security-header checks for production.
- Decide whether to keep WordPress as the public surface or graduate to MediaWiki/GitBook/hybrid architecture after production evidence accumulates.

## Done Means For This Production Pass

- `studio wp` can activate the theme and plugins.
- The home page and key routes render locally.
- Handoff files are complete enough for a new AI to onboard.
- Public copy clearly says LlmWikis is non-normative and UAIX is canonical for UAI-1.
- Route, lint, activation, and HTTP smoke checks pass or any blocker is recorded.
- Final responses name targeted checks run and clearly say when package/release checks were out of scope.

Why This File Exists

This is a UAI AI Memory handoff file from aiwikis.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 progress 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: LlmWikis Progress; Recently Completed; Current Focus; Known Blockers; Next Work; Done Means For This Production Pass. 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

Provenance And History

  • Current observation: 2026-05-02T01:47:31.8867765Z
  • Source origin: current-source-workspace
  • Retrieval method: local-source-workspace
  • Duplicate group: sfg-407 (primary)
  • Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "LlmWikis Progress",
    "source_site":  "aiwikis.org",
    "source_url":  "https://aiwikis.org/",
    "canonical_url":  "https://aiwikis.org/files/aiwikis/raw-system-archives-llmwikis-internal-memory-reorg-2026-05-01-uai-progre-e3fd6033/",
    "source_reference":  "raw/system-archives/llmwikis/internal-memory-reorg/2026-05-01/.uai/progress.uai",
    "file_type":  "uai",
    "content_category":  "uai-system",
    "content_hash":  "sha256:e3fd6033b818bc4066a61f9815a5ee6052202e75074d20ac60013d12a395f2c3",
    "last_fetched":  "2026-05-02T01:47:31.8867765Z",
    "last_changed":  "2026-04-29T13:35:18.0182259Z",
    "import_status":  "unchanged",
    "duplicate_group_id":  "sfg-407",
    "duplicate_role":  "primary",
    "related_files":  [

                      ],
    "generated_explanation":  true,
    "explanation_last_generated":  "2026-05-02T01:47:31.8867765Z"
}