LLMWikis Agent File Handoff

LlmWikis dogfoods chat-start file intake beside the UAIX Project Handoff pattern.

Metadata

Field	Value
Source site	llmwikis.org
Source URL	https://llmwikis.org/
Canonical AIWikis URL	https://aiwikis.org/llmwikis/files/uai-file-handoff-uai-2b6b9cff/
Source reference	.uai/file-handoff.uai
File type	uai
Content category	uai-system
Last fetched	2026-06-22T01:56:21.9510185Z
Last changed	2026-06-13T22:16:56.6058656Z
Content hash	sha256:2b6b9cff85604676ac192f3a24a8e67cad67bf87ac6e1f35e3937d1fa5f2745d
Import status	unchanged
Raw source layer	data/sources/llmwikis/uai-file-handoff-uai-2b6b9cff8560.uai
Normalized source layer	data/normalized/llmwikis/uai-file-handoff-uai-2b6b9cff8560.txt

Current File Content

Structure Preview

• LlmWikis Agent File Handoff
• Purpose
• Current Local Behavior
• Bucket Meaning
• Routes
• Publication Boundary
• 2026-06-13 Long-Run Agent Wiki Intake Outcome
• Required AI Behavior

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: operations
title: "LlmWikis Agent File Handoff"
created: "2026-04-27"
updated: "2026-06-13"
author: "LlmWikis maintainers"
version: 11
---

# LlmWikis Agent File Handoff

## Purpose

LlmWikis dogfoods chat-start file intake beside the UAIX Project Handoff pattern.

Humans and other AI systems can drop files into two intent buckets:

- `agent-file-handoff/Content/` for candidate public content, handbook drafts, report drafts, public source material, and editorial material that might become a site route after review.
- `agent-file-handoff/Improvement/` for audits, SEO reports, QA findings, bug notes, screenshots of broken UI, roadmap suggestions, and other work items for improving the site.

When LLM Wiki/AIWikis long memory is configured, processed originals must not move into a source-site `agent-file-handoff/Archive/` folder. After explicit disposition, completed work when safe and relevant, hot-memory outcome, long-memory preservation, and targeted-check reporting, preserve the original in the configured LLM Wiki/AIWikis evidence path with checksums and a manifest, then remove the source-site copy unless the human explicitly keeps that named file active.

Cross-site evidence preservation is the long-memory workflow. Processed originals are preserved in AIWikis.org under `[local path redacted]` or another named LLM Wiki evidence path, with source paths, final paths, checksums, dispositions, source-site removal status, and transfer evidence recorded. The AIWikis copy is memory/provenance evidence, not canonical LlmWikis public truth and not UAIX authority.

Hot-context reorganization follows the same evidence-first rule. Before slimming `AGENTS.md`, `readme.human`, or loaded `.uai` records, copy the pre-slim versions into AIWikis cold memory with source paths, final paths, byte counts, SHA-256 hashes, and a compiled source/concept/log summary. The 2026-05-01 reorganization uses `[local path redacted]`.

A new AI should scan both buckets while loading `AGENTS.md` and the listed `.uai` files, before broad planning. This keeps newly supplied files visible even when the human prompt does not mention them.

Visibility is not enough. If `.uai/intake-index.uai` lists any non-blocked file, the AI must inspect the file immediately, summarize it, and give it an explicit disposition before unrelated broad work continues.

Disposition is not enough either. For every safe, relevant file, the AI must extract at least one actionable slice and complete at least one named project-work action before source-site removal. Project work can be public copy, handbook docs, tests, roadmap/progress, decision or issue state, code, package metadata, discovery records, or another accepted LlmWikis system surface. Memory distribution without site or system work is a failed handoff unless every active file is unsafe, duplicate, out of scope, or truly blocked with a durable reason. Do not count copying a report into `.uai`, AIWikis, or an LLM Wiki as the project work by itself.

A complete intake outcome has five slots:

1. Reviewed summary/disposition.
2. Actual site or system work completed, or a durable blocker explaining why no safe work can be completed.
3. Hot-memory update or explicit no-change reason.
4. Long-memory preservation evidence when configured, or explicit `not configured`.
5. Checks run or skipped, with blockers.
6. Source-site removal or explicit keep-active reason.

The requirement is the AGENTS.md-triggered intake behavior. `scripts/Invoke-UaiFileIntake.ps1` is the local helper for refreshing the index. Watchers, daemons, queue folders, manifests, background services, cron loops, out-of-chat auto-pickup processes, and bucket-local README instructions are discouraged because the pattern must remain portable across languages and environments.

## Current Local Behavior

- Content bucket: `agent-file-handoff/Content/`
- Improvement bucket: `agent-file-handoff/Improvement/`
- Source-site Archive folder: forbidden by default when LLM Wiki/AIWikis long memory is configured.
- Downstream long-term evidence: `[local path redacted]`.
- Legacy compatibility folder: `agent-file-handoff/intake/`
- Chat-start index helper: `scripts/Invoke-UaiFileIntake.ps1`
- Generated index: `.uai/intake-index.uai`
- No watcher/daemon helpers, queue folders, manifests, or background services are part of this workspace's intake pattern.

The intake pass computes SHA-256 checksums, assigns bucket names and route hints, marks non-executable files as `needs-agent-review`, and updates the index. It does not move files, publish files, create review drafts, or mark dropped files trusted.

The legacy `agent-file-handoff/intake/` folder is retained only so old instructions have a place to point. Files left there are treated as `Unsorted` and should be moved, applied, deferred, clarified, or blocked before broad work continues.

Do not create a source-site Archive bucket for routine intake when LLM Wiki/AIWikis long memory is configured. Durable AI behavior belongs in `AGENTS.md` and `.uai` files, and human-facing explanation belongs in root `readme.human`.

The required reviewed disposition for every `needs-agent-review` file is one of:

- `apply-now` because it is directly relevant and safe.
- `convert-into-roadmap-progress` because it is useful but not public content or not ready for implementation.
- `preserve-as-duplicate` only when the same useful information has already been processed with work, hot-memory outcome, LLM Wiki/AIWikis preservation outcome, checks, and source-site removal or keep-active reason.
- `defer-with-reason` only when no safe actionable slice can be promoted now.
- `ask-for-clarification` when the file cannot be acted on safely without the human.
- `block-as-unsafe-or-out-of-scope` for unsafe, unsupported, out-of-scope, or executable/script-like files.

Use `defer-with-reason` narrowly. Do not defer a safe, relevant broad or future-facing report merely because it also needs later roadmap, source, or support-claim review; promote the current useful slice and record the rest as background when a long-memory path exists.

## Bucket Meaning

- `Content` means the file might become public-facing copy, a public page, a report, an example, or editorial source material after review.
- `Improvement` means the file is about fixing, auditing, prioritizing, or improving the site. It should become roadmap/progress state, an issue, a code change, a QA check, or an implementation note, not public copy by default.
- LLM Wiki/AIWikis evidence means the file already received a disposition and is preserved for provenance. Do not treat it as active intake unless a human explicitly names the preserved file or moves a copy back into `Content/` or `Improvement/`.
- `Unsorted` means the file was dropped into the legacy `intake/` folder and needs human or agent classification before promotion.

## Routes

- `site-content-draft` for text-like files in `Content/`.
- `site-improvement-report` for text-like files in `Improvement/`.
- `handoff-context` for `.uai`, `.json`, `.yaml`, and `.yml`.
- `content-asset-review` or `improvement-evidence-review` for PDFs and images, depending on bucket.
- `package-review` for `.zip`.
- `content-data-review` or `improvement-data-review` for `.csv` and `.tsv`, depending on bucket.
- `unsorted-review` for old files left in `intake/`.

Executable/script-like drops are reported as `blocked-local-review` in the index and must not be executed or promoted.

## Publication Boundary

The handoff contract is a one-shot intake scan and review gate during `AGENTS.md` load. A project can implement that scan in any language or perform it manually, as long as both `Content/` and `Improvement/` are inspected, no source-site Archive bucket is created when LLM Wiki/AIWikis long memory is configured, `.uai/intake-index.uai` is updated, and every listed file has a complete intake outcome before unrelated broad work begins.

Public LlmWikis page changes, WordPress writes, discovery files, release ZIPs, and production deployment still require the normal constraints and checks. UAIX.org remains canonical for UAI-1, Project Handoff, schemas, validator behavior, roadmap, governance, and changelog.

## 2026-06-13 Long-Run Agent Wiki Intake Outcome

- Processed Improvement reports: `Integrating HIPIF Hierarchical Planning into LLMWikis.md` and `Integrating HORMA Hierarchical Memory LLMWikis.md`.
- Disposition: both `apply-now` for safe, relevant implementation lessons about long-running agent work, folded state, active subgoals, completed summaries, bounded retrieval, staged synthesis, and process-quality checks.
- Actual work: promoted the accepted slice into `/guide/long-run-agent-wikis/`, setup wizard `?mode=long-run-agent`, generated setup model/packet fields, starter bundle templates/files, starter preflight rules, discovery files, and regression tests.
- Boundary: public copy intentionally uses neutral LlmWikis terms and does not expose source-report acronyms, external research/project names, optimizer/model/benchmark names, hardware/training/performance claims, hosted service claims, hidden-memory claims, public MCP/write API claims, autonomous publishing claims, certification, endorsement, or conformance claims.
- Preservation status: the prior source-site Archive copies were retired in favor of AIWikis long-memory evidence; `.uai/intake-index.uai` refreshed and now reports zero active intake files.

## Required AI Behavior

At the start of a broad task:

1. Read `AGENTS.md`.
2. Check `agent-file-handoff/Content/` and `agent-file-handoff/Improvement/`, then refresh the index. Use `scripts/Invoke-UaiFileIntake.ps1` when available.
3. Load the listed `.uai` files, including the refreshed `.uai/intake-index.uai`.
4. Inspect each `needs-agent-review` file named in the index before unrelated broad work.
5. Summarize each file and state one reviewed disposition.
6. For every safe, relevant file, complete at least one named site or system work item before source-site removal.
7. Update hot memory with the accepted current facts, changed surfaces, checks, blockers, and next work, or record an explicit no-change reason.
8. Preserve long-memory source material, summaries, evidence paths, checksums, dispositions, source-site removal status, and transfer evidence in AIWikis or another named LLM Wiki when configured; otherwise record `not configured`.
9. Remove processed source-site copies only after the work, durable records, and preservation evidence exist unless the human explicitly keeps a named file active.
10. Promote useful content into public pages, `.uai` files, roadmap state, or release evidence only after checking source status, UAIX authority boundaries, and review requirements.
11. Report back with three lists: intake files processed, information saved to hot memory versus long-term wiki evidence, and actual site/system work completed.

Why This File Exists

This is a UAI AI Memory handoff 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 .uai file is a compact operations 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 Agent File Handoff; Purpose; Current Local Behavior; Bucket Meaning; Routes; Publication Boundary; 2026-06-13 Long-Run Agent Wiki Intake Outcome; Required AI Behavior. 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-219 (primary)
• Historical hash records are stored in data/hashes/source-file-history.jsonl.

Machine-Readable Metadata

{
    "title":  "LLMWikis Agent File Handoff",
    "source_site":  "llmwikis.org",
    "source_url":  "https://llmwikis.org/",
    "canonical_url":  "https://aiwikis.org/llmwikis/files/uai-file-handoff-uai-2b6b9cff/",
    "source_reference":  ".uai/file-handoff.uai",
    "file_type":  "uai",
    "content_category":  "uai-system",
    "content_hash":  "sha256:2b6b9cff85604676ac192f3a24a8e67cad67bf87ac6e1f35e3937d1fa5f2745d",
    "last_fetched":  "2026-06-22T01:56:21.9510185Z",
    "last_changed":  "2026-06-13T22:16:56.6058656Z",
    "import_status":  "unchanged",
    "duplicate_group_id":  "sfg-219",
    "duplicate_role":  "primary",
    "related_files":  [

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