Hub Page Excellence Implementation Plan

For agentic workers: REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (- [ ]) syntax for tracking.

Goal: Bring docs/index.md and docs/water/index.md to 22/22 on the hub-page excellence rubric (4 lenses), codify the rubric into STANDARDS.md § Hub Pages, and queue 11 follow-up retrofit tasks for the other Foundation hubs.

Architecture: Three-phase pipeline. Phase 1: draft rubric → Codex critiques completeness → reconcile. Phase 2: per-page scorecard → edit plan → Codex per-lens critique → reconcile → implement (run once for home, once for water). Phase 3: promote rubric to STANDARDS.md → Codex enforceability critique → reconcile → queue retrofit tasks. Final commit via publisher per Survipedia convention.

Tech Stack: Markdown (MkDocs Material), mcp__codex__codex MCP server (gpt-5.5 independent critic), scripts/category-coverage.py, mkdocs build --strict, Survipedia state files (LOG.md, MEMORY-*, AUDIT_LOG.md, BACKLOG.md, JOURNAL.md).

Spec: docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md

Working artifacts directory: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/ — committed alongside final work. Contains rubric versions and Codex raw responses. Scorecards and edit plans are ephemeral (live in conversation, summarized in commit body).

Pre-flight

Task 0: Verify environment

Files: none

Step 1: Confirm Codex MCP is wired

Run: grep -A3 '"codex"' /root/projects/wiki/.mcp.json Expected: codex server entry present.

Step 2: Confirm working directory is clean for hub-excellence work

Run: git status --short Expected: only the spec file + plan file showing (or clean). If unrelated changes exist, stash them first.

Step 3: Load deferred tool schema for the Codex MCP

Use ToolSearch with query select:mcp__codex__codex to load the schema before any Phase 1 call.

Step 4: Create the working-artifacts directory

Run: mkdir -p /root/projects/wiki/docs/superpowers/specs/2026-05-14-hub-page-excellence-working Expected: directory exists.

Step 5: Append start-of-loop entry to LOG.md

Append to /root/projects/wiki/LOG.md:

YYYY-MM-DD HH:MM | webmaster | hub-excellence:start | spec + plan loaded, phase 1 entering | ~1k tokens

Use actual UTC timestamp.

Phase 1: Rubric anchor

Task 1: Persist RUBRIC.v1 from the spec

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/rubric-v1.md

Step 1: Write RUBRIC.v1 as a standalone artifact

Write the file with the exact 22 items from the spec's "The rubric — RUBRIC.v1 (22 items, 4 lenses)" section. Preserve the IDs (R1–R6, A1–A5, S1–S5, C1–C6). No prose preamble — bullet list only.

Step 2: Sanity-check item count

Run: grep -cE '^- \*\*[RASC][0-9]+\*\*' /root/projects/wiki/docs/superpowers/specs/2026-05-14-hub-page-excellence-working/rubric-v1.md Expected: 22

Task 2: Codex Phase 1 — rubric completeness

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/codex-phase1-response.md

Step 1: Send Phase 1 prompt to Codex

Use mcp__codex__codex with the Phase 1 prompt template from the spec verbatim. Paste the full RUBRIC.v1 file contents at the end. Set Codex's model to its default (gpt-5.5).

Step 2: Save Codex's raw response

Write Codex's full response (verbatim, no editing) to codex-phase1-response.md with a header:

# Codex Phase 1 Response — Rubric Completeness
**Date:** YYYY-MM-DD
**Prompt sent:** RUBRIC.v1 (22 items) + Phase 1 template

---
<paste verbatim response>

Step 3: Log the round-trip

Append to LOG.md:

YYYY-MM-DD HH:MM | codex-mcp | hub-excellence:phase-1 | rubric-v1 completeness critique | ~Nk tokens

Estimate tokens from response length.

Task 3: Reconcile to RUBRIC.v2

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/rubric-v2.md

Step 1: Apply the expansion guard

Read Codex's response. For each "missing item" Codex flagged, decide accept/reject using the spec's guard: - Accept only if the item closes a clear coverage gap in one of the 4 lenses - Reject if it's stylistic preference or merely a re-phrasing of an existing item - Modify an existing item only if Codex flagged genuine ambiguity

For each "redundant item" Codex flagged, decide whether to merge or keep separate.

Step 2: Write RUBRIC.v2

Same structure as RUBRIC.v1. Add a reconciliation footer:

---
## Reconciliation log (RUBRIC.v1 → v2)
- [Codex flag: "missing X"] accept; added as <new ID>
- [Codex flag: "R3 and S2 overlap"] reject; reasoning: R3 is reader-question framing, S2 is search-passage hierarchy; same surface, different intent
- ...

Step 3: Verify final item count

Run: grep -cE '^- \*\*[RASC][0-9]+\*\*' /root/projects/wiki/docs/superpowers/specs/2026-05-14-hub-page-excellence-working/rubric-v2.md Expected: 18–28 (started at 22; guard allows modest expansion).

Step 4: Append phase-1 close entry to LOG.md

YYYY-MM-DD HH:MM | webmaster | hub-excellence:phase-1-close | rubric-v2 finalized, N items | ~2k tokens

Phase 2a: Home page edit pass

Task 4: Score `docs/index.md` against RUBRIC.v2

Files: none persisted; scorecard lives in conversation

Step 1: Read the current homepage

Read /root/projects/wiki/docs/index.md in full.

Step 2: Score every rubric item

For each item in RUBRIC.v2, write one line in the conversation: - R1: pass — first 80 words establish "encyclopedia of practical self-reliance" + 12 Foundations + 3 domains. No jargon. - R4: fail — no Field Note admonition present. Has tip/onramp but not a labeled Field Note. - (etc. for all items)

Each line must justify the pass/fail with a specific page reference.

Step 3: Tally fails

Count pass / partial / fail. This is the scorecard. Goal after edits: 22+ passes (or all hard-fails passing + soft-fails documented).

Task 5: Draft the home-page edit plan

Files: none persisted; edit plan lives in conversation

Step 1: Group fails by lens

Group every fail/partial from Task 4 by lens (R/A/S/C).

Step 2: Propose one edit per fail

For each fail, write:

[edit ID] scope: <word|sentence|section|new-block>
  current: "<exact text or 'absent'>"
  proposed: "<exact replacement>"
  rationale: closes <rubric ID> by <one line>

Sacred-cow guard: do NOT propose edits that remove the hero image, the 3-step onramp tip, the four scenario tabs, the tier framework (Essentials/Infrastructure/Fieldcraft), or the Field Note pattern (if absent, ADD one; do not remove other admonitions to "make room").

Step 3: Quick sanity check on tone

Read the proposed edits as a block. Verify: calm, practical, empowering. No fearmongering. No political framing. No "you must" pressure.

Task 6: Codex Phase 2 — home page per-lens critique

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/codex-phase2-home.md

Step 1: Send Phase 2 prompt to Codex

Use the Phase 2 prompt template from the spec verbatim. Bundle into the prompt: 1. RUBRIC.v2 (full) 2. Current docs/index.md (full, with line numbers if possible) 3. The edit plan from Task 5 (full)

Step 2: Save Codex's raw response

Write to codex-phase2-home.md with header similar to Phase 1.

Step 3: Log the round-trip

Append to LOG.md:

YYYY-MM-DD HH:MM | codex-mcp | hub-excellence:phase-2-home | per-lens critique of docs/index.md | ~Nk tokens

Task 7: Reconcile Codex critique → final home edit plan

Files: none persisted; final edit plan lives in conversation, summarized in commit body

Step 1: Walk each Codex critique item

For every flag in codex-phase2-home.md, decide one of: - accept — add to the edit plan or modify an existing edit - reject — write a one-line reason (e.g., "rejected: violates Q4 sacred cow on tier structure") - modify — accept the spirit but change scope or wording

Webmaster arbitrates. Decisions must be defensible against the rubric, not stylistic preference.

Step 2: Produce the final edit plan

Same format as Task 5 plus a reconciliation block:

## Codex reconciliation
  - [Codex: "intro buries the 'why now'"] accept; redraft paragraph 1
  - [Codex: "drop scenario tabs, they fragment SEO"] reject; reason: Q4 sacred cow ("Preserve current voice + Field Note + tier structure" leaves tabs in scope, but reader segmentation closes C2 — keep)
  - ...

Task 8: Implement home page edits

Files: - Modify: docs/index.md

Step 1: Apply edits one lens at a time

Use Edit tool for each item in the final edit plan. Apply clarity edits first, then authority, then SEO, then conversion. This order minimizes rework — clarity changes can invalidate SEO edits otherwise.

Step 2: Hero image rule check

Run: ls -la /root/projects/wiki/docs/assets/images/shared/gl_home-hero.webp Expected: file exists, > 0 bytes. If missing, dispatch @image-gen via Agent tool with flux-2-max (landscape 16:9, text-free).

Step 3: Rebuild and rescore

Walk through every rubric item in RUBRIC.v2 against the new file content. Confirm 22/22 (or document waivers with rationale in the commit body).

Step 4: Update date_modified in front matter

The homepage front matter currently lacks date_modified. Add it under content_type: utility:

date_modified: 2026-05-14

Use today's date.

Step 5: Append phase-2a close entry to LOG.md

YYYY-MM-DD HH:MM | webmaster | hub-excellence:phase-2a-close | docs/index.md scored 22/22 (or N/22 + waivers) | ~Xk tokens

Phase 2b: Water hub edit pass

Task 9: Score `docs/water/index.md` against RUBRIC.v2

Files: none persisted

Step 1: Read the current water hub

Read /root/projects/wiki/docs/water/index.md in full.

Step 2: Score every rubric item

Same format as Task 4. Note water already has a labeled Field Note (!!! tip "Field note") and a purification decision-tree diagram. Likely to score higher on R4 and A5 than home did.

Step 3: Tally fails

Task 10: Draft the water-hub edit plan

Files: none persisted

Step 1: Group fails by lens, propose edits

Same format as Task 5. Sacred-cow guard: preserve the three-layers framing (Sourcing / Purification / Storage), the hero image, the decision-tree diagram, the existing Field Note, and the "Where to start" checklist pattern.

Step 2: Sanity-check tone

Verify no fearmongering. The current page already has !!! warning "Dehydration kills faster than hunger" — keep this; it's calibrated warning, not panic.

Task 11: Codex Phase 2 — water hub per-lens critique

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/codex-phase2-water.md

Step 1: Send Phase 2 prompt to Codex with RUBRIC.v2 + current water hub + Task 10 edit plan.
Step 2: Save Codex's raw response to codex-phase2-water.md.
Step 3: Log the round-trip

YYYY-MM-DD HH:MM | codex-mcp | hub-excellence:phase-2-water | per-lens critique of docs/water/index.md | ~Nk tokens

Task 12: Reconcile Codex critique → final water edit plan

Files: none persisted

Step 1: Walk each Codex critique item — accept / reject / modify with reasoning. Same arbitration rules as Task 7.
Step 2: Produce the final water edit plan with reconciliation block.

Task 13: Implement water hub edits

Files: - Modify: docs/water/index.md

Step 1: Apply edits — same lens order as Task 8 (clarity → authority → SEO → conversion).
Step 2: Hero image rule check

Run: ls -la /root/projects/wiki/docs/assets/images/water/gl_water-hub-hero.webp Expected: file exists, text-free. (Will be visually verified in Task 17.)

Step 3: Update date_modified

The water hub currently has no date_modified (and lacks page_id, foundation, date_published, depth, criticality — but per STANDARDS.md Section 5 the hub index.md exemption applies; only child pages need those). Add date_modified: 2026-05-14 to the front matter.

Step 4: Re-score against RUBRIC.v2

Confirm 22/22 or document waivers.

Step 5: Update Guide Sync Metadata footer

The water hub has a Guide Sync footer (lines 57-62). Update Last Major Update to today's date and verify Affects Guides: [first-30-days] still applies. Add any other affected guides discovered during the rewrite.

Step 6: Append phase-2b close entry to LOG.md

YYYY-MM-DD HH:MM | webmaster | hub-excellence:phase-2b-close | docs/water/index.md scored 22/22 (or N/22 + waivers) | ~Xk tokens

Phase 3: Codify into STANDARDS.md

Task 14: Draft STANDARDS.md § Hub Pages

Files: - Modify: STANDARDS.md (append new section; do NOT modify existing sections in this task)

Step 1: Read STANDARDS.md to find the right insertion point

Run: grep -n '^## ' /root/projects/wiki/STANDARDS.md Identify the last top-level section. The Hub Pages section is appended after it (or before any "Appendix" if present).

Step 2: Draft the section

The section MUST contain (per spec):

## Hub Page Excellence Standard

**Applies to:** `docs/index.md` (homepage) and every Foundation hub `docs/{foundation}/index.md` (12 total).
**Enforced by:** editor (scorecard); publisher (hard-fails block commit).
**Origin:** 2026-05-14 brainstorm (docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md).

Hub pages carry the majority of Foundation entry traffic. They are the storefront. They must hold up to a hostile reader and to search-engine scrutiny equally. Excellence on a hub page means satisfying every item in the 22-item (or final v2 count) rubric across four lenses: reader clarity, authority signal, SEO, and conversion to deeper reading.

### The rubric

[Insert RUBRIC.v2 verbatim, with each item followed by a one-line "pass example from <home or water>" and one-line "fail example from pre-rewrite version".]

### Anti-patterns

- Bare link lists without "why click" lines (violates C3)
- Generic encyclopedia intros that don't establish stakes (violates R1)
- H2 sections that label taxonomy ("The three layers") rather than answer reader questions (violates R3)
- Closing sections that summarize instead of giving a next action (violates C4)
- "You must" pressure language or fearmongering (violates Operating Constitution rule 6)

### Hard-fail items (publisher blocks commit)

- **A4** life-safety treatment per SOURCES.md
- **R4** ≥1 Field Note admonition present
- **S1** description ≤160 chars

### Soft-fail items (editor flags, does not block)

All other rubric items. Editor documents waived items in the commit body with a rationale.

### Editor checklist

- [ ] Score the page against every rubric item
- [ ] Confirm zero hard-fails
- [ ] Document any soft-fail waivers with rationale
- [ ] Verify hero image present and text-free
- [ ] Verify `date_modified` updated

### Retrofit protocol

The 11 remaining Foundation hubs are retrofitted via `HUB-RETROFIT:` BACKLOG tasks (one per hub, `[M]` 4pt). Retrofit tasks are `POST-SPRINT:` tagged and deferred until 2026-06-16 or until sprint mode ends — they fall under the "OTHER" tier of the sprint priority ladder.

Step 3: Use the Edit tool to append the new section to STANDARDS.md.

Task 15: Codex Phase 3 — enforceability critique

Files: - Create: docs/superpowers/specs/2026-05-14-hub-page-excellence-working/codex-phase3-standards.md

Step 1: Send Phase 3 prompt to Codex with the new STANDARDS.md section (just the new section, not the whole file).
Step 2: Save Codex's raw response to codex-phase3-standards.md.
Step 3: Log the round-trip

YYYY-MM-DD HH:MM | codex-mcp | hub-excellence:phase-3 | STANDARDS.md § Hub Pages enforceability critique | ~Nk tokens

Task 16: Reconcile and refine STANDARDS.md

Files: - Modify: STANDARDS.md (the new Hub Pages section)

Step 1: Walk each Codex enforceability flag

For each "ambiguous item" Codex flagged, either reword for testability or accept the ambiguity with a one-line editor note. For each "missing enforcement detail," add the detail.

Step 2: Re-edit STANDARDS.md with the reconciled changes.
Step 3: Append rationale to AUDIT_LOG.md (governance edit per Operating Constitution rule 5)

Append to AUDIT_LOG.md:

## 2026-05-14 — STANDARDS.md § Hub Pages added

- **What changed:** New top-level section "Hub Page Excellence Standard" appended.
- **Why:** Codifies the rubric developed during the 2026-05-14 hub-excellence brainstorm; makes the standard enforceable across all 12 Foundation hubs + homepage.
- **Triggered by:** Approved brainstorm spec `docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md`.
- **Reviewed by:** Codex (gpt-5.5) for enforceability via `codex-phase3-standards.md`.
- **Reversibility:** Section is self-contained; revert by removing the section.

Phase 4: Backlog + memory + commit

Task 17: Queue 11 HUB-RETROFIT tasks in BACKLOG.md

Files: - Modify: BACKLOG.md

Step 1: List the 11 remaining hubs

The 12 Foundations are: water, food, medical, shelter, energy, security, community, threats, skills, tools (gear), mobility, mindset. Water is done in this loop, leaving 11.

Step 2: Append one BACKLOG task per remaining hub

Use the Survipedia task format. Example for food:

- [ ] [M] T3 P3 HUB-RETROFIT: POST-SPRINT: Apply Hub Pages Excellence Standard to docs/food/index.md — score against rubric, edit to 22/22, document waivers. Source: 2026-05-14 hub-excellence brainstorm.
  RICE: <reach>×<impact>÷<effort> = <score>

  <details><summary>Learning Artifact</summary>

  signal: docs/food/index.md not yet scored against STANDARDS.md § Hub Pages rubric
  source: docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md
  source_tier: 2
  detected_by: webmaster (hub-excellence pilot)
  date: 2026-05-14
  affected_pages: [docs/food/index.md]
  confidence: high
  recommended_action: After 2026-06-16 sprint end, run the 3-phase hub-excellence pipeline on this page (no Phase 1 needed — RUBRIC.v2 already exists).
  state_destination: BACKLOG.md
  escalate_human: no

  </details>

Repeat for: medical, shelter, energy, security, community, threats, skills, tools, mobility, mindset. Each task is its own block (do not abbreviate "similar to food" — the engineer may read out of order).

RICE scoring: use traffic-weighted reach where possible (medical, food, water are typically highest-reach Foundations; mindset, mobility lower). Use 1 for impact (qualitative improvement), 4 for effort ([M] task).

Step 3: Verify task count

Run: grep -c '^- \[ \] \[M\] T[0-9]* P[0-9]* HUB-RETROFIT:' /root/projects/wiki/BACKLOG.md Expected: at least 11 (may be more if pre-existing retrofit tasks exist; the new batch adds 11).

Task 18: Update memory topic files

Files: - Modify: MEMORY-CONTENT.md - Modify: MEMORY-STANDARDS.md

Step 1: Append to MEMORY-CONTENT.md

Add a section under the appropriate heading (or create a new one):

## Hub page excellence (2026-05-14)

What made docs/index.md and docs/water/index.md excellent after the 2026-05-14 rewrite:

- **Clarity per lens:** First 80 words establish stakes and topic without jargon. H2 sections answer reader questions, not topic taxonomy.
- **Authority signals:** Multi-setting breadth (urban/suburban/rural) and named methods (not just descriptions) — readers see depth at a glance.
- **SEO win pattern:** Featured-snippet candidates (S5) sit near the top in a complete <50-word answer; ≥8 outgoing child-page links distribute hub authority.
- **Conversion lever:** Every child link gets a one-line "why click" — no bare list items. Closing section ends with one next action, not a summary.

Apply when authoring or retrofitting other Foundation hubs. Full rubric: STANDARDS.md § Hub Pages.

Step 2: Append to MEMORY-STANDARDS.md

## Hub Page Excellence Standard

**Pointer:** STANDARDS.md § Hub Pages (added 2026-05-14).

Every hub page (homepage + 12 Foundation hubs) scored against a 22-item (or v2 final count) rubric covering reader clarity, authority, SEO, and conversion. Editor enforces; publisher blocks hard-fails (A4, R4, S1).

When editing any hub page, score before/after against the rubric. Document any waived soft-fails in the commit body.

Task 19: Update MEMORY.md index

Files: - Modify: MEMORY.md

Step 1: Add index pointers

Append two lines to MEMORY.md (after existing index entries, before the 200-line truncation point):

- [Hub page excellence](MEMORY-CONTENT.md) — what made home/water hubs score 22/22 against the rubric (2026-05-14)
- [Hub Page Excellence Standard](MEMORY-STANDARDS.md) — pointer to STANDARDS.md § Hub Pages; rubric, hard-fails, retrofit protocol

Task 20: Reflector pass — JOURNAL.md + close TODOs

Files: - Modify: JOURNAL.md - Modify: any in-flight BACKLOG.md task representing this work (if one exists)

Step 1: Append a JOURNAL.md narrative entry

Use the existing JOURNAL.md format. Cover: - What was done (hub-excellence pilot on home + water, STANDARDS.md § Hub Pages added) - What worked (Codex's independent critique on each lens) - What surprised (any reconciliations that changed your mind; any Codex flags that were genuinely valuable vs noise) - What to apply next (retrofit pattern, lessons for hubs 3–13)

Step 2: Check BACKLOG.md for an in-flight task representing this work

Run: grep -i 'hub.*excellence\|hub.*perfect\|home.*water.*rubric' /root/projects/wiki/BACKLOG.md If a task exists, mark it - [x] complete.

Step 3: Append LOOP_COMPLETE to LOG.md

YYYY-MM-DD HH:MM | reflector | LOOP_COMPLETE | hub-excellence pilot: home + water shipped, STANDARDS.md § Hub Pages added, 11 retrofit tasks queued | ~Yk tokens

Task 21: Pre-publish verification

Files: none modified; verification only

Step 1: Run mkdocs build --strict

Run: cd /root/projects/wiki && mkdocs build --strict 2>&1 | tail -40 Expected: INFO - Documentation built in ... seconds with no warnings about broken links, missing nav entries, or schema errors. If any warning appears, fix it before commit.

Step 2: Confirm rubric scorecards on both pages

In conversation: state each page's final score (e.g., "home: 22/22 — 0 waivers" or "water: 21/22 — C2 waived: reason"). This becomes input to the commit body.

Step 3: Confirm hero rule on both pages

Run:

ls -la /root/projects/wiki/docs/assets/images/shared/gl_home-hero.webp /root/projects/wiki/docs/assets/images/water/gl_water-hub-hero.webp

Expected: both exist, > 0 bytes.

If either is missing or you have any reason to suspect text-leak, dispatch Skill: visual-review on the file before commit.

Step 4: Confirm all state files have entries

Run:

grep -c '2026-05-14.*hub-excellence' /root/projects/wiki/LOG.md

Expected: ≥ 8 entries (start, phase-1-codex, phase-1-close, phase-2-home-codex, phase-2a-close, phase-2-water-codex, phase-2b-close, phase-3-codex, LOOP_COMPLETE — could be more).

Also confirm:

grep -l '2026-05-14' /root/projects/wiki/AUDIT_LOG.md /root/projects/wiki/JOURNAL.md /root/projects/wiki/MEMORY-CONTENT.md /root/projects/wiki/MEMORY-STANDARDS.md /root/projects/wiki/MEMORY.md

Expected: every file listed. If one is missing the 2026-05-14 string, that file wasn't updated — return to the task that handles it.

Task 22: Publisher commit + push

Files: all modified files staged

Step 1: Invoke the publisher skill

Use Skill: publisher. The publisher runs its own validations (offline-sufficiency check on canon-cell pages, etc.), stages with git add -A, commits, and pushes.

If publisher is unavailable in this session, fall back to manual:

Step 2: Manual fallback — stage all changes

Run: cd /root/projects/wiki && git add -A && git status --short Verify the staged file list matches expectations: - docs/index.md - docs/water/index.md - STANDARDS.md - AUDIT_LOG.md - BACKLOG.md - LOG.md - JOURNAL.md - MEMORY.md, MEMORY-CONTENT.md, MEMORY-STANDARDS.md - docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md - docs/superpowers/plans/2026-05-14-hub-page-excellence.md - docs/superpowers/specs/2026-05-14-hub-page-excellence-working/*.md (rubric-v1, rubric-v2, codex-phase1, codex-phase2-home, codex-phase2-water, codex-phase3-standards)

No other files should be staged. If any unrelated file appears, investigate before committing.

Step 3: Manual fallback — commit

cd /root/projects/wiki && git commit -m "$(cat <<'EOF'
feat: hub-page excellence pilot — home + water rewritten, STANDARDS.md § Hub Pages added

Pilot of the hub-page excellence pipeline. Brainstormed 2026-05-14
(docs/superpowers/specs/2026-05-14-hub-page-excellence-design.md), executed via
3-phase workflow with Codex (gpt-5.5) as independent per-lens critic.

Results:
- docs/index.md: scored N/22 → 22/22 (or N/22 + documented waivers)
- docs/water/index.md: scored N/22 → 22/22 (or N/22 + documented waivers)
- STANDARDS.md § Hub Pages: 22-item rubric, hard-fails (A4/R4/S1), retrofit protocol
- BACKLOG.md: 11 HUB-RETROFIT POST-SPRINT tasks queued (one per remaining Foundation hub)
- AUDIT_LOG.md: governance edit rationale recorded
- MEMORY-CONTENT.md / MEMORY-STANDARDS.md: lessons + standard pointer

Codex round-trips: 4 (phase-1 rubric completeness, phase-2 home, phase-2 water, phase-3 standards enforceability). Reconciliation logged inline.

Sprint freeze compliance: existing-page rewrite, not new pages. Retrofit tasks tagged POST-SPRINT — defer until 2026-06-16.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
EOF
)"

Step 4: Push

Run: cd /root/projects/wiki && git push origin main Expected: push succeeds. GitHub Actions deploys automatically on push.

Step 5: Confirm deploy

The GitHub Actions deploy runs unobserved. The local commit landing + push success is the close-of-loop signal.

Self-review checklist (run this AFTER writing the plan, BEFORE handoff)

This is for the plan author, not the engineer.

Every spec section maps to at least one task (architecture → Phase 1–4 tasks; rubric → Task 1, 3, 14; codex protocol → Tasks 2, 6, 11, 15; per-page edit format → Tasks 5, 7, 10, 12; STANDARDS.md → Task 14; state files → Tasks 17–20; verification gates → Task 21)
No "TBD" / "TODO" / "similar to" / "appropriate" placeholders
Codex prompt templates appear verbatim in the relevant tasks or by reference to the spec (spec is committed, so reference is OK)
File paths absolute and exact
Naming consistent across tasks (RUBRIC.v1 vs rubric-v1.md, etc.)
Sprint freeze and Operating Constitution rule 5 honored (audit log entry for STANDARDS.md edit)
Hero image rule check appears for both pages

Hub Page Excellence Implementation Plan

Pre-flight

Task 0: Verify environment

Phase 1: Rubric anchor

Task 1: Persist RUBRIC.v1 from the spec

Task 2: Codex Phase 1 — rubric completeness

Task 3: Reconcile to RUBRIC.v2

Phase 2a: Home page edit pass

Task 4: Score docs/index.md against RUBRIC.v2

Task 5: Draft the home-page edit plan

Task 6: Codex Phase 2 — home page per-lens critique

Task 7: Reconcile Codex critique → final home edit plan

Task 8: Implement home page edits

Phase 2b: Water hub edit pass

Task 9: Score docs/water/index.md against RUBRIC.v2

Task 10: Draft the water-hub edit plan

Task 11: Codex Phase 2 — water hub per-lens critique

Task 12: Reconcile Codex critique → final water edit plan

Task 13: Implement water hub edits

Phase 3: Codify into STANDARDS.md

Task 14: Draft STANDARDS.md § Hub Pages

Task 15: Codex Phase 3 — enforceability critique

Task 16: Reconcile and refine STANDARDS.md

Phase 4: Backlog + memory + commit

Task 17: Queue 11 HUB-RETROFIT tasks in BACKLOG.md

Task 18: Update memory topic files

Task 19: Update MEMORY.md index

Task 20: Reflector pass — JOURNAL.md + close TODOs

Task 21: Pre-publish verification

Task 22: Publisher commit + push

Self-review checklist (run this AFTER writing the plan, BEFORE handoff)

Task 4: Score `docs/index.md` against RUBRIC.v2

Task 9: Score `docs/water/index.md` against RUBRIC.v2