yzr-llm-wiki-management

• SKILL.md (this document) — workflows + boundaries
• Deterministic execution is handled by llmw CLI — this skill is zero-code, converged into two commands; refer to the detector for inconsistencies:

  • llmw check-fixtures

    : Only detect (output drift report, no disk writes)

  • llmw upgrade

    : Workspace skeleton + per-wiki aggregated deterministic upgrade (dry-run by default)

• scan → Write

  <workspace>/INDEX.md

  +

  <workspace>/STATS.md

  (format A2/A3)
• query → Provide answers in conversation (with per-wiki references); optional archive to

  <workspace>/cross_queries/<slug>.md

  (format A4)
• link → Add cross-wiki links in involved wikis via the

  yzr-llm-wiki-management

  ingest process
• lint → Write

  <workspace>/LINT.md

  (format A5) + conversation summary
• upgrade → Run

  llmw upgrade

  (dry-run by default; add

  --apply [--yes]

  to write to disk); 3 final-state JSON outputs from CLI; see §6 for details

1. Locate workspace path:

   $LLMW_WORKSPACE

   → default

   ~/yzr-llm-wiki-workspace

   → interactive query
2. Verify that

   <workspace>/workspace.toml

   exists — if not, prompt the user "Workspace not initialized yet, run

   llmw init

   to initialize" (do not run it for the user)
3. Load cross-wiki MEMORY index: Automatically loaded via

   @MEMORY/MEMORY.md

   import in

   <workspace>/AGENTS.md

   when working in the workspace root; if working in a non-root directory / reading AGENTS.md natively without expanding

   @

   agents → explicitly

   Read <$LLMW_WORKSPACE>/MEMORY/MEMORY.md

   to supplement
4. Load scope boundaries: When the agent's cwd is within a

   <wiki>/

   subdirectory and switches to running

   yzr-llm-wiki-management

   , this skill's discipline does not take over; instead, the single-wiki discipline in

   <wiki>/AGENTS.md

   takes effect
5. Do not run

   scan

   automatically — wait for user's operation intent

1. Read

   <workspace>/workspace.toml

   to get the

   [wikis]

   registry
2. For each wiki:
   • Read

     <wiki>/wiki_metadata.toml

     (maintained by CLI)
   • Read the "Boundaries of This Wiki" section in

     <wiki>/AGENTS.md

     (get boundaries)
   • Read

     <wiki>/wiki/index.md

     (existing content + paragraph skeleton)
   • Scan

     <wiki>/wiki/{entities,concepts,sources,comparisons,syntheses}/

     to get page counts
   • Recursively scan

     <wiki>/raw/

     to count raw materials (only count, do not read content)
   • Read the last entry in

     <wiki>/wiki/log.md

     to get last activity
   • Read

     <wiki>/MEMORY/

     to count memory files (only file names)
3. Read the

   <workspace>/MEMORY/MEMORY.md

   index, aggregate according to A2 sorting rules, write INDEX.md + STATS.md (format A2/A3)
4. Atomic write (POSIX

   tmp + fsync + rename

   )
5. Report in conversation: "INDEX.md / STATS.md refreshed, X wikis, Y pages, Z raw materials"

yzr-llm-wiki-management

• If the answer involves a single wiki →

  <wiki>/wiki/syntheses/<slug>.md

  (via

  yzr-llm-wiki-management

  )
• If the answer involves multiple wikis →

  <workspace>/cross_queries/<slug>.md

  (directly written by this skill, format A4)

yzr-llm-wiki-management

references/ingest-workflow.md

1. Scan: For each wiki's

   wiki/entities/

   +

   wiki/concepts/

   , extract all entity names (frontmatter

   title

   or filename slug)
2. Deduplicate and aggregate: Collect cross-wiki entities with the same name / similar meaning (compared by description) as candidate pairs
3. Suggest: List candidate pairs in conversation and let the user select which ones need cross-wiki links
4. Write: After user confirmation, for each involved wiki, call the ingest process of

   yzr-llm-wiki-management

   to update the corresponding entity / concept page — append a "Cross-wiki References" section (xref format see the "Cross-wiki Agreements" section of

   <workspace>/AGENTS.md

   )

1. Workspace-level deterministic checks (inline in agent):
   • Cross-wiki duplicate entities (pairs with same name but different slugs)
   • Invalid cross-wiki links (paths in

     sources

     of cross_queries/*.md do not exist;

     ../<another-wiki>/...

     paths in

     <wiki>/wiki/**

     do not exist)
   • Unregistered wiki subdirectories (

     <wiki>/AGENTS.md

     exists on disk but not registered in workspace.toml)
   • Wikis registered in workspace.toml but not present on disk (orphaned registrations)
   • Consistency between wiki lists in STATS.md and INDEX.md
   • MEMORY index consistency: Scan

     <workspace>/MEMORY/*.md

     (excluding

     MEMORY.md

     ), if any file is not listed in

     MEMORY/MEMORY.md

     index → report

     memory-not-indexed

     (severity = info)
2. Semi-qualitative checks:
   • Whether wikis with overlapping topics need to be merged
   • Whether the tag system is chaotic (same tag with different meanings / same meaning with different tag names)
3. What this skill does not do: Single-wiki internal lint — delegate to

   yzr-llm-wiki-management

4. Output: Write

   <workspace>/LINT.md

   (format A5) + report in conversation

yzr-llm-wiki-management

cross_queries/

<workspace>/AGENTS.md

1. Identify observations worth preserving → self-check scope to confirm cross-wiki
2. Determine entry format (full / short)
3. Write to

   MEMORY/<slug>.md

   (full entry) or append a short entry line directly to

   MEMORY/MEMORY.md

4. Synchronize one line in MEMORY.md index (missing write = unreadable next time, lint

   memory-not-indexed

   as fallback)

<workspace>/INDEX.md

STATS.md

LINT.md

<wiki>/MEMORY/

llmw upgrade

<workspace>/AGENTS.md

llmw check-fixtures

1. llmw upgrade

   defaults to dry-run → output workspace + processing plan for each wiki + 3 final-state JSON (add

   --json

   for machine-readable)
2. Interpret 3 final states:

   • done

     → Wrap up, prompt user "Workspace upgrade completed, X wikis upgraded"

   • blocked_drift

     → Move custom content to

     MEMORY/

     according to the

     hint

     field, then re-run

     llmw upgrade --apply --yes

   • verify_failed

     → Report to user, escalate to manual processing
3. Subsequent content migration for each wiki uses the

   yzr-llm-wiki-management

   workflow — this skill does not run it on behalf of the user

INDEX.md

STATS.md

LINT.md

User: "What records do I have about RAID across all my wikis?"

1. Skill reads

   <workspace>/INDEX.md

   → finds that

   huawei_storage_wiki

   's description contains "storage"
2. Mode = synthesis
3. Delegates query "RAID" to

   yzr-llm-wiki-management

   for

   huawei_storage_wiki

4. Gets answer (with source page references), provides it to user in conversation, and asks "This only involves 1 wiki, should we archive it to

   huawei_storage_wiki/wiki/syntheses/raid-overview.md

   ?"
5. User confirms → uses

   yzr-llm-wiki-management

   to write synthesis page + log entry

User: "I just downloaded an LLM inference paper, which wiki should I put it in?"

1. Skill reads INDEX.md + reads description / tags of each wiki
2. Mode = route
3. Returns: "

   huawei_storage_wiki

   focuses on storage, not relevant;

   test

   wiki focuses on test, not relevant; recommend creating a new wiki (

   llmw wiki --name=llm-inference add ...

   )"
4. Offers to run

   llmw wiki --name=llm-inference add

   on behalf of the user — execute after user confirmation (skill handles routing and closes the loop)

• Required Reading:

  <workspace>/AGENTS.md

  (= rendered from

  workspace-agents-md-template.md

  template, byte-owned) — canonical workspace-level contract
• Single-wiki Contract:

  yzr-llm-wiki-management

  SKILL.md + references/ (this skill reads wiki files according to its contract, does not write directly)
• CLI Documentation: Workspace CLI (command

  llmw

  , maintained in the same repo as this skill) — reference for commands like

  init / add / remove / config / enter / model ...

  here

This appendix carries format contracts for files written to disk by the skill. Reading contracts and instance ownership can be found in

<workspace>/AGENTS.md

.

• Path:

  <workspace-root>/workspace.toml

• Complete schema authority is in CLI code;

  workspace-toml-reads-satisfied

  check verifies reading contract integrity

scan

upgrade

templates_version

llmw upgrade

workspace_format

wiki_format

[wikis.<name>].path

scan

[wikis.<name>].created_at

• Maintainer: Skill writes during

  scan

  ; CLI does not write
• Required frontmatter (A7) +

  type: workspace-index

  ;

  title

  recommended as

  "Workspace Index"

  ;

  tags

  recommended as

  [workspace, index]

• Body skeleton:
  markdown

  # <Workspace Display Name> — Workspace Index
  
  > Workspace entry document. One section per wiki, sorted alphabetically by wiki name; sorted by `created_at` ascending within the same alphabetical group.
  
  ## Wikis
  
  ### <wiki-name>
  
  - **display_name**: ...
  - **topic**: ...
  - **description**: ...
  - **tags**: [...]
  - **created**: YYYY-MM-DD
  - **last activity**: YYYY-MM-DD (log entry kind)
  - **page counts**: 0 entities / 0 concepts / ...
  - **key entities**: [...]
  - **one-line summary**: ...
  
  ## Cross-wiki Links
  ...(short description)
  ## Recent Activity (across all wikis)
  ...

• Maintainer: Skill writes alongside

  scan

  ; difference from INDEX.md: structured (table)
• Required frontmatter (A7) +

  type: workspace-stats

• Body skeleton:

  # <Workspace> — Workspace Stats

  +

  ## Overview

  summary table +

  ## Per-wiki

  section with per-wiki sub-tables (pages / entities / concepts / sources / comparisons / syntheses / raw_files / last_log_entry / tags / memory_files)
• Skill write scenario:

  scan

  (same refresh as INDEX.md)

• Maintainer: Skill writes when query output is suitable for archiving
• File naming:

  <slug>.md

  , kebab-case

  ^[a-z0-9][a-z0-9-]*$

  (A8)
• Required frontmatter (A7) +

  type: cross-query

  ;

  tags

  recommended as

  [workspace, cross-query, <tags of involved wikis>...]

  ; required

  sources

  (array of paths to wiki pages referenced, relative to workspace root) +

  wikis

  (array of involved wiki names)
• Skill write scenario: When user confirms archiving of query output

• Maintainer: Skill writes the latest report during

  lint

  (does not accumulate, overwrites each time, is a snapshot)
• Required frontmatter (A7) +

  type: workspace-lint

• Body skeleton:

  # <Workspace> — Lint Report (<YYYY-MM-DD>)

  +

  ## Per-wiki Issues

  (one section per wiki, single-wiki lint goes to yzr-llm-wiki-management) +

  ## Workspace-level Issues

  (cross-wiki duplicate entities / unregistered subdirectories / outdated STATS / MEMORY index consistency / ...)
• Skill write scenario:

  lint

  (overwrites each time)

• Maintainer: CLI creates empty directory and writes

  MEMORY/MEMORY.md

  index placeholder during init; subsequent entries are written by skill + synchronized to append one line to MEMORY.md index. Humans do not write; CLI does not write
• MEMORY is not mandatory to be listed in INDEX.md (private agent entry)
• Entry format (full / short) + when to write/not write + index line format canonical = "Memory Discipline" section (byte-owned template rendering) of

  <workspace>/AGENTS.md

  , not repeated in this appendix