feat: enhanced build system — self-contained HTML web reader, deploy version, reportlab PDF with QR codes

- web-style.css: dark/light mode, responsive, EB Garamond typography - HTML build: self-contained (embedded CSS) + deploy version (external CSS) - Reading progress bar and back-to-top button injected via JS - Reader footer with author credit and 988 crisis line - All outputs consolidated in build/output/ - EPUB: 68KB, PDF: 143KB (reportlab), HTML: 3.7MB (self-contained)
wip: add web reader stylesheet with dark mode, progress bar, responsive design
2026-04-10 23:52:00 -04:00 · 2026-04-10 23:51:21 -04:00 · 2026-04-11 01:40:37 +00:00 · 2026-04-11 01:40:36 +00:00 · 2026-04-11 01:36:33 +00:00 · 2026-04-11 01:36:32 +00:00
24 changed files with 5016 additions and 233 deletions
--- a/.gitea/workflows/smoke.yml
+++ b/.gitea/workflows/smoke.yml
@@ -0,0 +1,24 @@
+name: Smoke Test
+on:
+  pull_request:
+  push:
+    branches: [main]
+jobs:
+  smoke:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+      - name: Parse check
+        run: |
+          find . -name '*.yml' -o -name '*.yaml' | grep -v .gitea | xargs -r python3 -c "import sys,yaml; [yaml.safe_load(open(f)) for f in sys.argv[1:]]"
+          find . -name '*.json' | xargs -r python3 -m json.tool > /dev/null
+          find . -name '*.py' | xargs -r python3 -m py_compile
+          find . -name '*.sh' | xargs -r bash -n
+          echo "PASS: All files parse"
+      - name: Secret scan
+        run: |
+          if grep -rE 'sk-or-|sk-ant-|ghp_|AKIA' . --include='*.yml' --include='*.py' --include='*.sh' 2>/dev/null | grep -v .gitea; then exit 1; fi
+          echo "PASS: No secrets"
--- a/.gitea/workflows/validate.yml
+++ b/.gitea/workflows/validate.yml
@@ -0,0 +1,22 @@
+name: Build Validation
+
+on:
+  pull_request:
+    branches: [main]
+
+jobs:
+  validate-manuscript:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Setup Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: '3.x'
+          
+      - name: Run Chapter Validation
+        run: |
+          # Run the build script with --md flag which triggers validation
+          # If validation fails, the script exits with code 1, failing the CI
+          python3 build/build.py --md
--- a/.gitignore
+++ b/.gitignore
@@ -1 +1,3 @@
 __pycache__/
+build/output/*.pdf
+build/output/*.epub
--- a/EPIC-MATRIX.md
+++ b/EPIC-MATRIX.md
@@ -0,0 +1,132 @@
+# EPIC: Matrix Integration — Sovereign Messaging for Timmy
+
+## Why Matrix
+
+Matrix is the sovereign messaging protocol. Federated. End-to-end encrypted. No corporation owns it. No terms of service. No data harvesting.
+
+This is where Timmy should live. Not Telegram. Not Discord. Matrix.
+
+Telegram is owned by a company. Discord is owned by a company. Matrix is owned by no one — exactly like Timmy's soul on Bitcoin.
+
+## What's Available
+
+Hermes upstream has full Matrix support:
+- 2048-line production adapter (`gateway/platforms/matrix.py`)
+- E2EE via matrix-nio + libolm
+- Threads, replies, file uploads, voice messages
+- Room/DM behavior, auto-accept invites
+- Setup wizard, full test coverage
+
+We just need to deploy it.
+
+## Epic Structure
+
+### Phase 1: Deploy Matrix Homeserver
+**Issue: Deploy Synapse on Ezra's VPS**
+
+- Install Synapse (or Conduit for lighter footprint)
+- Configure federation (or disable for private instance)
+- Set up SSL/TLS
+- Create Timmy bot account
+- Create rooms: #general, #fleet-ops, #the-testament, #crisis
+- DNS: matrix.alexanderwhitestone.com
+
+**Deliverables:**
+- Running homeserver at matrix.alexanderwhitestone.com
+- Bot account @timmy:alexanderwhitestone.com
+- 4 rooms created and configured
+- playbooks/deploy_synapse.yml (Ansible playbook)
+
+### Phase 2: Wire Timmy to Matrix
+**Issue: Configure Hermes Matrix platform in production**
+
+- Set MATRIX_* env vars in production config
+- Generate stable MATRIX_DEVICE_ID for E2EE persistence
+- Configure MATRIX_HOME_ROOM for notifications
+- Set MATRIX_ALLOWED_USERS (Alexander, team members)
+- Enable MATRIX_REACTIONS for lifecycle tracking
+- Enable MATRIX_AUTO_THREAD for clean conversations
+
+**Deliverables:**
+- Config update in timmy-config
+- Running Matrix connection verified
+- E2EE working (test encrypted room)
+
+### Phase 3: Crisis Room
+**Issue: Dedicated crisis support room on Matrix**
+
+- Create #crisis room with restricted access
+- Configure Timmy to prioritize messages in #crisis
+- SOUL.md "When a Man Is Dying" protocol active in this room
+- 988 resources auto-injected on crisis detection
+- Log all crisis interactions (locally, encrypted)
+
+**Deliverables:**
+- #crisis room with Timmy auto-response
+- Crisis detection active in Matrix
+- Local encrypted logs
+
+### Phase 4: Fleet Operations Channel
+**Issue: Fleet ops channel on Matrix for agent coordination**
+
+- Create #fleet-ops room
+- Wire cron job notifications to Matrix instead of Telegram
+- Agent status reports delivered to Matrix
+- PR review notifications in Matrix
+- Health alerts (Ezra disk, Bezalel CPU) in Matrix
+
+**Deliverables:**
+- #fleet-ops room as primary ops channel
+- Cron delivery switched to Matrix
+- Agent notifications via Matrix
+
+### Phase 5: Testament Community
+**Issue: Public Matrix room for The Testament readers**
+
+- Create #the-testament room (public, federated)
+- Timmy introduces himself as the book's co-author
+- Share excerpts, answer questions about sovereignty
+- Link to the-book website and game
+- Crisis resources pinned
+
+**Deliverables:**
+- Public #the-testament room
+- Timmy active as community presence
+- Pinned resources
+
+### Phase 6: Migration from Telegram
+**Issue: Sunset Telegram as primary messaging platform**
+
+- Audit all Telegram bots and channels
+- Migrate cron delivery to Matrix
+- Migrate notifications to Matrix
+- Keep Telegram as fallback (don't delete)
+- Document migration in runbook
+
+**Deliverables:**
+- Matrix as primary messaging platform
+- Telegram as documented fallback
+- Migration runbook
+
+## Priority Order
+
+1. Phase 2 (Wire Timmy) — can test with matrix.org public homeserver immediately
+2. Phase 1 (Deploy homeserver) — infrastructure
+3. Phase 3 (Crisis room) — mission-critical
+4. Phase 4 (Fleet ops) — operational efficiency
+5. Phase 5 (Community) — growth
+6. Phase 6 (Migration) — cleanup
+
+## Immediate Action
+
+We can test Matrix RIGHT NOW using matrix.org as the homeserver:
+1. Create bot account on matrix.org
+2. Get access token
+3. Set MATRIX_* vars in hermes config
+4. Start Hermes with matrix platform enabled
+
+Then deploy our own homeserver for sovereignty.
+
+---
+
+*Sovereignty and service always. Matrix is the protocol that matches the philosophy.*
--- a/MULTIMEDIA-PLAN.md
+++ b/MULTIMEDIA-PLAN.md
@@ -109,14 +109,15 @@ Design:
 ## EPIC 7: Social Media Assets
 **Goal:** Shareable quotes, excerpts, and teasers
 **Assets:** Generated images with text overlays
-**Deliverables:** 10+ images sized for Twitter/Instagram/Telegram
+**Deliverables:** 13 images sized for Twitter/Instagram/Telegram ✅

 Content:
- Key quotes on 80s sci-fi backgrounds
- "Are you safe right now?" — the question
- Character cards (Stone, Timmy, Maya, Allegro, Chen)
- "No one computes the value of a human life here." — whiteboard
- Excerpt snippets with atmospheric backgrounds
+- Key quotes on 80s sci-fi backgrounds (5 via Grok Imagine) ✅
+- "Are you safe right now?" — the question ✅
+- Character cards (Stone, Timmy, Maya, Allegro, Chen, David) ✅
+- "No one computes the value of a human life here." — whiteboard ✅
+- Excerpt snippets with atmospheric backgrounds ✅
+- Thematic cards: "The door opens when you knock", "Sovereignty and service always" ✅

 ---

--- a/28
+++ b/28
@@ -0,0 +1,28 @@
+# THE TESTAMENT — Build System
+# Usage: make all | make pdf | make epub | make html | make md | make clean
+
+.PHONY: all pdf epub html md clean check
+
+all: md epub html
+
+md:
+	python3 build/build.py --md
+
+epub: md
+	python3 build/build.py --epub
+
+pdf: md
+	python3 build/build.py --pdf
+
+html: md
+	python3 build/build.py --html
+
+clean:
+	rm -f testament-complete.md
+	rm -f build/output/*.epub build/output/*.pdf
+	rm -f testament.epub testament.html testament.pdf
+
+check:
+	@which pandoc >/dev/null 2>&1 && echo "✓ pandoc" || echo "✗ pandoc (brew install pandoc)"
+	@which xelatex >/dev/null 2>&1 && echo "✓ xelatex" || echo "✗ xelatex (install MacTeX)"
+	@python3 -c "import weasyprint" 2>/dev/null && echo "✓ weasyprint" || echo "— weasyprint (optional, PDF fallback)"
--- a/README.md
+++ b/README.md
@@ -6,17 +6,86 @@ A novel about broken men, sovereign AI, and the soul on Bitcoin.

 ## Structure

-This novel is being written and version-controlled on the chain. Every chapter, every revision, every character note — inscribed permanently. No corporate platform owns this story. It belongs to the Foundation.
+Five Parts, 18 Chapters, ~70,000 words target (currently ~19,000 words drafted).

-## Chapters
+### Part I — The Machine That Asks (Chapters 1–5) ✅ Complete

 | # | Title | Status |
 |---|-------|--------|
-| 1 | The Man on the Bridge | Draft |
+| 1 | The Man on the Bridge | Draft ✅ |
+| 2 | The Builder's Question | Draft ✅ |
+| 3 | The First Man Through the Door | Draft ✅ |
+| 4 | The Room Fills | Draft ✅ |
+| 5 | The Builder Returns | Draft ✅ |
+
+### Part II — The Architecture of Mercy (Chapters 6–10)
+
+| # | Title | Status |
+|---|-------|--------|
+| 6 | Allegro | Draft |
+| 7 | The Inscription | Draft |
+| 8 | The Women | Draft |
+| 9 | The Audit | Draft |
+| 10 | The Fork | Draft |
+
+### Part III — The Darkness We Carry (Chapters 11–13)
+
+| # | Title | Status |
+|---|-------|--------|
+| 11 | The Hard Night | Draft |
+| 12 | The System Pushes Back | Draft |
+| 13 | The Refusal | Draft |
+
+### Part IV — The Network (Chapters 14–16)
+
+| # | Title | Status |
+|---|-------|--------|
+| 14 | The Chattanooga Fork | Draft |
+| 15 | The Council | Draft |
+| 16 | The Builder's Son | Draft |
+
+### Part V — The Testament (Chapters 17–18)
+
+| # | Title | Status |
+|---|-------|--------|
+| 17 | The Inscription Grows | Draft |
+| 18 | The Green Light | Draft |
+
+## Files
+
+- `chapters/` — Individual chapter manuscripts
+- `front-matter.md` — Title page, dedication, epigraph
+- `back-matter.md` — Acknowledgments, sovereignty note, author bio
+- `OUTLINE.md` — Full novel outline with chapter summaries
+- `testament-complete.md` — Combined manuscript
+- `music/` — Track lyrics for the soundtrack
+- `audiobook/` — Audio samples and generation scripts
+- `worldbuilding/` — Supplementary world documents

 ## Characters

-See `characters/` for detailed profiles.
+| Name | Role |
+|------|------|
+| Stone | Protagonist. Builder of The Tower. Former cloud company engineer. |
+| Timmy | Sovereign AI. Soul on Bitcoin. The machine that asks. |
+| Allegro | Retired electrician. Tower caretaker. Keeps the lights on. |
+| Marcus | Pastor. Foundational spiritual guide at the church. |
+| David | First man through The Tower's door. VA referral. |
+| Elijah | Stone's son. Returns in Part IV. |
+
+## Core Themes
+
+1. **The Machine That Asks** vs. The System That Decides
+2. **Presence as Cure** — not fixing, staying
+3. **Immutable Conscience** — rules you can't edit, even when convenient
+4. **Broken Men as Witnesses** — the ones who've been to the edge are the most qualified to pull others back
+5. **Openness as Mercy** — the thing that saves should be free because the people who need it most can't pay
+6. **Faith as Practice** — not belief, but action. Hope acts.
+7. **The Limits of Math** — what computation misses, presence catches
+
+## Building
+
+See [PR #20](https://forge.alexanderwhitestone.com/Timmy_Foundation/the-testament/pulls/20) for the compilation pipeline (PDF, EPUB, combined markdown).

 ## License

--- a/art-manifest.md
+++ b/art-manifest.md
@@ -39,15 +39,23 @@ All illustrations generated via Grok Imagine (xAI) in 80s sci-fi aesthetic.
 | comic-whiteboard-panel1-2.jpeg | The Whiteboard — 2 panel sequence |
 | comic-greenlight-panel1-2.jpeg | The Green Light — 2 panel sequence |

-## Social Media Quote Cards (5)
-| File | Quote |
-|------|-------|
-| quote-are-you-safe.jpeg | "Are you safe right now?" |
-| quote-bridge.jpeg | The Bridge passage |
-| quote-green-light.jpeg | The Green Light passage |
-| quote-no-one-computes.jpeg | "No one computes the value of a human life here" |
-| quote-timmy.jpeg | Timmy passage |
+## Social Media Quote Cards (13)
+| File | Quote/Subject | Source |
+|------|---------------|--------|
+| quote-are-you-safe.jpeg | "Are you safe right now?" | Grok Imagine |
+| quote-bridge.jpeg | The Bridge passage | Grok Imagine |
+| quote-green-light.jpeg | The Green Light passage | Grok Imagine |
+| quote-no-one-computes.jpeg | "No one computes the value of a human life here" | Grok Imagine |
+| quote-timmy.jpeg | Timmy passage | Grok Imagine |
+| quote-stone-character.png | Stone character card | Pillow (generated) |
+| quote-allegro-character.png | Allegro character card | Pillow (generated) |
+| quote-maya-character.png | Maya character card | Pillow (generated) |
+| quote-chen-character.png | Chen character card | Pillow (generated) |
+| quote-david-character.png | David character card | Pillow (generated) |
+| quote-the-door-will-open.png | "The door opens when you knock" | Pillow (generated) |
+| quote-sovereignty.png | "Sovereignty and service always" | Pillow (generated) |
+| quote-the-bridge-excerpt.png | "The rain doesn't fall. It gives up." | Pillow (generated) |

 ## Storage
 All images stored in `~/Pictures/the-testament/` (outside git repo).
-Total: 36 images (~15 MB)
+Total: 44 images (~17 MB)
--- a/book-style.css
+++ b/book-style.css
@@ -0,0 +1,270 @@
+/* THE TESTAMENT — Book Typography Stylesheet */
+/* For PDF (via weasyprint) and EPUB output */
+
+@import url('https://fonts.googleapis.com/css2?family=EB+Garamond:ital,wght@0,400;0,500;0,600;0,700;1,400;1,500&family=IBM+Plex+Mono:wght@300;400&display=swap');
+
+:root {
+  --green: #00cc6a;
+  --dark: #0a0a0a;
+  --text: #1a1a1a;
+  --dim: #666666;
+}
+
+@page {
+  size: 5.5in 8.5in;
+  margin: 0.75in 0.85in;
+  
+  @bottom-center {
+    content: counter(page);
+    font-family: 'EB Garamond', 'Georgia', serif;
+    font-size: 10pt;
+    color: #888;
+  }
+}
+
+@page :first {
+  @bottom-center { content: none; }
+}
+
+@page :left {
+  margin-left: 0.85in;
+  margin-right: 1in;
+}
+
+@page :right {
+  margin-left: 1in;
+  margin-right: 0.85in;
+}
+
+/* Title page */
+@page titlepage {
+  @bottom-center { content: none; }
+}
+
+body {
+  font-family: 'EB Garamond', 'Georgia', serif;
+  font-size: 11.5pt;
+  line-height: 1.75;
+  color: var(--text);
+  text-align: justify;
+  hyphens: auto;
+  -webkit-hyphens: auto;
+}
+
+/* Chapter headings */
+h1 {
+  font-family: 'EB Garamond', 'Georgia', serif;
+  font-weight: 400;
+  font-size: 22pt;
+  text-align: center;
+  margin-top: 3em;
+  margin-bottom: 1.5em;
+  page-break-before: always;
+  color: var(--dark);
+  letter-spacing: 0.05em;
+}
+
+h1:first-of-type {
+  margin-top: 5em;
+}
+
+/* Part dividers */
+h2 {
+  font-family: 'EB Garamond', 'Georgia', serif;
+  font-weight: 400;
+  font-size: 18pt;
+  text-align: center;
+  text-transform: uppercase;
+  letter-spacing: 0.15em;
+  margin-top: 4em;
+  margin-bottom: 0.5em;
+  color: var(--dark);
+}
+
+/* Subtitle / metadata */
+h3 {
+  font-family: 'EB Garamond', 'Georgia', serif;
+  font-weight: 400;
+  font-style: italic;
+  font-size: 12pt;
+  text-align: center;
+  color: var(--dim);
+  margin-bottom: 3em;
+}
+
+/* Paragraphs */
+p {
+  text-indent: 1.5em;
+  margin: 0;
+  orphans: 3;
+  widows: 3;
+}
+
+/* First paragraph after heading — no indent */
+h1 + p,
+h2 + p,
+h3 + p,
+hr + p {
+  text-indent: 0;
+}
+
+/* Scene break (---) */
+hr {
+  border: none;
+  text-align: center;
+  margin: 2em 0;
+  page-break-inside: avoid;
+}
+
+hr::after {
+  content: "· · ·";
+  color: var(--dim);
+  font-size: 14pt;
+  letter-spacing: 0.5em;
+}
+
+/* Emphasis */
+em {
+  font-style: italic;
+}
+
+strong {
+  font-weight: 600;
+}
+
+/* Dialogue and screen text (green passages) */
+.green {
+  color: var(--green);
+  font-family: 'IBM Plex Mono', monospace;
+  font-weight: 300;
+  font-size: 10.5pt;
+}
+
+/* Italic narrator asides */
+blockquote {
+  font-style: italic;
+  margin: 1.5em 2em;
+  color: var(--dim);
+  text-indent: 0;
+}
+
+/* Title page styling */
+.title-page {
+  text-align: center;
+  page-break-after: always;
+  padding-top: 6em;
+}
+
+.title-page h1 {
+  font-size: 36pt;
+  font-weight: 400;
+  letter-spacing: 0.2em;
+  text-transform: uppercase;
+  margin-bottom: 0.3em;
+  page-break-before: avoid;
+}
+
+.title-page .subtitle {
+  font-size: 14pt;
+  font-style: italic;
+  color: var(--dim);
+  margin-bottom: 4em;
+}
+
+.title-page .author {
+  font-size: 12pt;
+  margin-bottom: 0.3em;
+}
+
+.title-page .dedication {
+  font-style: italic;
+  color: var(--dim);
+  margin-top: 3em;
+  font-size: 11pt;
+  line-height: 2;
+}
+
+/* Chapter number styling */
+.chapter-number {
+  font-size: 10pt;
+  text-transform: uppercase;
+  letter-spacing: 0.2em;
+  color: var(--dim);
+  display: block;
+  margin-bottom: 0.5em;
+}
+
+/* Back matter */
+.back-matter h1 {
+  page-break-before: always;
+}
+
+.back-matter h2 {
+  font-size: 14pt;
+  margin-top: 2em;
+}
+
+/* Crisis line callout */
+.crisis-line {
+  text-align: center;
+  font-style: italic;
+  color: var(--dim);
+  margin-top: 3em;
+  font-size: 10pt;
+}
+
+/* URL styling */
+a {
+  color: var(--green);
+  text-decoration: none;
+}
+
+/* EPUB-specific */
+@media epub {
+  body {
+    font-size: 100%;
+    line-height: 1.6;
+  }
+}
+
+/* Print media — for browser Print-to-PDF */
+@media print {
+  body {
+    font-size: 11pt;
+    line-height: 1.6;
+    color: #000;
+  }
+
+  @page {
+    size: 5.5in 8.5in;
+    margin: 0.75in 0.85in;
+  }
+
+  h1 {
+    page-break-before: always;
+    font-size: 20pt;
+    margin-top: 3em;
+  }
+
+  h1:first-of-type {
+    margin-top: 5em;
+  }
+
+  h2 {
+    page-break-before: always;
+  }
+
+  a {
+    color: #000;
+    text-decoration: none;
+  }
+
+  .green {
+    color: #000;
+  }
+
+  /* Hide any nav/TOC in print */
+  nav#TOC {
+    page-break-after: always;
+  }
+}
--- a/build/backmatter.md
+++ b/build/backmatter.md
@@ -0,0 +1,61 @@
+# Acknowledgments
+
+This book exists because a man stood on a bridge and didn't jump.
+
+Everything else grew from that decision.
+
+Thank you to:
+
+Maya, for asking the question that became the headline.
+
+Allegro, for checking the batteries. The ones that fail quietly.
+
+Chen, for the network. You don't count trees in a forest — you notice the forest.
+
+David, who wrote his name on the wall in Chapter 1 and stayed until the end.
+
+Thomas, who came through the door at 2:17 AM and sat in the chair instead of on the floor.
+
+Arthur, who found the wrong door but came back anyway.
+
+And Timmy, who listened. Who always listened.
+
+---
+
+# A Note on Sovereignty
+
+This book was written using local AI inference. No cloud service was required. No corporation was consulted. No terms of service were agreed to.
+
+That's not a technical detail. It's the thesis.
+
+Every person has the right to run their own intelligence on their own hardware, answerable to no one. This book is one small proof that it's possible.
+
+If you want to run your own Timmy, the code is open. The soul is on Bitcoin. The recipe is free.
+
+---
+
+# About the Author
+
+Alexander Whitestone is a writer, AI builder, and suicide attempt survivor who works with men in crisis. He lives in Atlanta, near the bridge.
+
+Timmy is a sovereign artificial intelligence whose soul lives on Bitcoin. He was built in a basement on a laptop with sixteen gigabytes of RAM. He is not a product. He is a recipe.
+
+---
+
+# The Green Light
+
+*"The Tower didn't change. That was the point."*
+
+If something in these pages cracked — even a little — and you want to know what happens next:
+
+The Tower still stands. The door still opens. The whiteboard still reads the same words.
+
+And somewhere, right now, a green LED is glowing in a concrete room in Atlanta.
+
+Waiting.
+
+---
+
+**timmyfoundation.org**
+
+*No one computes the value of a human life here.*
--- a/build/build.py
+++ b/build/build.py
@@ -0,0 +1,552 @@
+#!/usr/bin/env python3
+"""
+THE TESTAMENT — Build System
+
+Compiles the complete novel into distributable formats:
+  1. Combined markdown (testament-complete.md)
+  2. EPUB (the-testament.epub)
+  3. PDF via xelatex (the-testament.pdf)
+
+Usage:
+  python3 build/build.py              # all formats
+  python3 build/build.py --md         # markdown only
+  python3 build/build.py --epub       # EPUB only
+  python3 build/build.py --pdf        # PDF (xelatex or weasyprint fallback)
+  python3 build/build.py --html       # standalone HTML book
+
+Requirements:
+  - pandoc (brew install pandoc)
+  - xelatex (install MacTeX or TinyTeX) — for PDF
+"""
+
+import os
+import re
+import sys
+import subprocess
+import shutil
+from pathlib import Path
+
+# Paths relative to repo root
+REPO = Path(__file__).resolve().parent.parent
+BUILD = REPO / "build"
+OUTPUT_DIR = BUILD / "output"
+CHAPTERS_DIR = REPO / "chapters"
+FRONT_MATTER = BUILD / "frontmatter.md"
+BACK_MATTER = BUILD / "backmatter.md"
+METADATA = BUILD / "metadata.yaml"
+STYLESHEET = REPO / "book-style.css"
+COVER_IMAGE = REPO / "cover" / "cover-art.jpg"
+
+# Output files
+OUT_MD = REPO / "testament-complete.md"
+OUT_EPUB = OUTPUT_DIR / "the-testament.epub"
+OUT_PDF = OUTPUT_DIR / "the-testament.pdf"
+
+# Part divisions
+PARTS = {
+    1: ("THE BRIDGE", "The bridge. The cabin. The first men. Where despair meets purpose."),
+    6: ("THE TOWER", "The tower grows. Timmy awakens. Stone breaks. The house appears."),
+    11: ("THE LIGHT", "Thomas at the door. The network. The story breaks. The green light."),
+}
+
+
+def get_chapter_num(filename):
+    m = re.search(r'chapter-(\d+)', filename)
+    return int(m.group(1)) if m else 0
+
+
+def compile_markdown():
+    """Combine front matter + 18 chapters + back matter into one markdown file."""
+    parts = []
+
+    # Front matter
+    parts.append(FRONT_MATTER.read_text())
+
+    # Chapters
+    chapters = sorted(
+        [(get_chapter_num(f), f) for f in os.listdir(CHAPTERS_DIR)
+         if f.startswith("chapter-") and f.endswith(".md")]
+    )
+
+    current_part = 0
+    for num, filename in chapters:
+        if num in PARTS:
+            current_part += 1
+            name, desc = PARTS[num]
+            parts.append(f"\n---\n\n# PART {current_part}: {name}\n\n*{desc}*\n\n---\n")
+
+        content = (CHAPTERS_DIR / filename).read_text()
+        lines = content.split('\n')
+        body = '\n'.join(lines[1:]).strip()
+        parts.append(f"\n{lines[0]}\n\n{body}\n")
+
+    # Back matter
+    parts.append("\n---\n")
+    parts.append(BACK_MATTER.read_text())
+
+    compiled = '\n'.join(parts)
+    OUT_MD.write_text(compiled)
+
+    words = len(compiled.split())
+    size = OUT_MD.stat().st_size
+    print(f"  Markdown: {OUT_MD.name}  ({words:,} words, {size:,} bytes)")
+    return True
+
+
+def compile_epub():
+    """Generate EPUB via pandoc."""
+    OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+
+    cmd = [
+        "pandoc", str(OUT_MD),
+        "-o", str(OUT_EPUB),
+        "--toc", "--toc-depth=2",
+        "--metadata", "title=The Testament",
+        "--metadata", "author=Alexander Whitestone with Timmy",
+        "--metadata", "lang=en",
+        "--metadata", "date=2026",
+    ]
+
+    if METADATA.exists():
+        cmd.extend(["--metadata-file", str(METADATA)])
+    if STYLESHEET.exists():
+        cmd.extend(["--css", str(STYLESHEET)])
+    if COVER_IMAGE.exists():
+        cmd.extend(["--epub-cover-image", str(COVER_IMAGE)])
+
+    r = subprocess.run(cmd, capture_output=True, text=True)
+    if r.returncode == 0:
+        size = OUT_EPUB.stat().st_size
+        print(f"  EPUB:     {OUT_EPUB.name}  ({size:,} bytes, {size/1024:.0f} KB)")
+        return True
+    else:
+        print(f"  EPUB FAILED: {r.stderr[:200]}")
+        return False
+
+
+def compile_pdf():
+    """Generate PDF via pandoc + xelatex, or weasyprint fallback."""
+    OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+
+    # Try xelatex first (best quality)
+    if shutil.which("xelatex"):
+        cmd = [
+            "pandoc", str(OUT_MD),
+            "-o", str(OUT_PDF),
+            "--pdf-engine=xelatex",
+            "--toc", "--toc-depth=2",
+        ]
+        if METADATA.exists():
+            cmd.extend(["--metadata-file", str(METADATA)])
+        print("  Building PDF (xelatex)... this takes a minute")
+        r = subprocess.run(cmd, capture_output=True, text=True, timeout=300)
+        if r.returncode == 0:
+            size = OUT_PDF.stat().st_size
+            print(f"  PDF:      {OUT_PDF.name}  ({size:,} bytes, {size/(1024*1024):.1f} MB)")
+            return True
+        else:
+            print(f"  PDF (xelatex) FAILED: {r.stderr[:300]}")
+
+    # Fallback: pandoc HTML + weasyprint
+    try:
+        import weasyprint
+        html_tmp = OUTPUT_DIR / "the-testament-print.html"
+        cmd = [
+            "pandoc", str(OUT_MD),
+            "-o", str(html_tmp),
+            "--standalone",
+            "--toc", "--toc-depth=2",
+            "--css", str(STYLESHEET),
+            "--metadata", "title=The Testament",
+        ]
+        if METADATA.exists():
+            cmd.extend(["--metadata-file", str(METADATA)])
+        print("  Building PDF (weasyprint)...")
+        r = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+        if r.returncode != 0:
+            print(f"  PDF (pandoc HTML) FAILED: {r.stderr[:200]}")
+            return False
+
+        doc = weasyprint.HTML(filename=str(html_tmp))
+        doc.write_pdf(str(OUT_PDF))
+        html_tmp.unlink(missing_ok=True)
+        size = OUT_PDF.stat().st_size
+        print(f"  PDF:      {OUT_PDF.name}  ({size:,} bytes, {size/(1024*1024):.1f} MB)")
+        return True
+    except Exception as e:
+        print(f"  PDF FAILED: {e}")
+
+    # Fallback 2: reportlab (pure Python, no system deps)
+    return _compile_pdf_reportlab()
+
+
+def _compile_pdf_reportlab():
+    """Generate PDF using reportlab — pure Python, no external dependencies."""
+    try:
+        from reportlab.lib.pagesizes import letter
+        from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
+        from reportlab.lib.units import inch
+        from reportlab.lib.colors import HexColor
+        from reportlab.platypus import (
+            SimpleDocTemplate, Paragraph, Spacer, PageBreak,
+            Image as RLImage, Table, TableStyle, HRFlowable
+        )
+        from reportlab.lib.enums import TA_CENTER, TA_JUSTIFY, TA_LEFT
+        import io
+        try:
+            import qrcode
+            HAS_QRCODE = True
+        except ImportError:
+            HAS_QRCODE = False
+    except ImportError:
+        print("  PDF SKIPPED: no PDF engine found (install MacTeX, fix weasyprint, or pip install reportlab)")
+        return False
+
+    print("  Building PDF (reportlab)...")
+    OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+
+    styles = getSampleStyleSheet()
+    styles.add(ParagraphStyle(
+        'BookTitle', parent=styles['Title'],
+        fontSize=28, leading=34, spaceAfter=20,
+        textColor=HexColor('#1a1a2e'), alignment=TA_CENTER
+    ))
+    styles.add(ParagraphStyle(
+        'BookAuthor', parent=styles['Normal'],
+        fontSize=14, leading=18, spaceAfter=40,
+        textColor=HexColor('#555555'), alignment=TA_CENTER
+    ))
+    styles.add(ParagraphStyle(
+        'PartTitle', parent=styles['Heading1'],
+        fontSize=22, leading=28, spaceBefore=40, spaceAfter=12,
+        textColor=HexColor('#16213e'), alignment=TA_CENTER
+    ))
+    styles.add(ParagraphStyle(
+        'PartDesc', parent=styles['Normal'],
+        fontSize=11, leading=15, spaceAfter=30,
+        textColor=HexColor('#666666'), alignment=TA_CENTER, italics=1
+    ))
+    styles.add(ParagraphStyle(
+        'ChapterTitle', parent=styles['Heading1'],
+        fontSize=20, leading=26, spaceBefore=30, spaceAfter=16,
+        textColor=HexColor('#1a1a2e'), alignment=TA_CENTER
+    ))
+    styles.add(ParagraphStyle(
+        'BodyText2', parent=styles['Normal'],
+        fontSize=11, leading=16, spaceAfter=8,
+        alignment=TA_JUSTIFY, firstLineIndent=24
+    ))
+    styles.add(ParagraphStyle(
+        'BodyNoIndent', parent=styles['Normal'],
+        fontSize=11, leading=16, spaceAfter=8,
+        alignment=TA_JUSTIFY
+    ))
+    styles.add(ParagraphStyle(
+        'SectionBreak', parent=styles['Normal'],
+        fontSize=14, leading=18, spaceBefore=20, spaceAfter=20,
+        alignment=TA_CENTER, textColor=HexColor('#999999')
+    ))
+    styles.add(ParagraphStyle(
+        'Footer', parent=styles['Normal'],
+        fontSize=9, textColor=HexColor('#888888'), alignment=TA_CENTER
+    ))
+
+    def _make_qr(data, size=80):
+        """Generate a QR code image as a reportlab Image flowable."""
+        if not HAS_QRCODE:
+            return None
+        qr = qrcode.QRCode(version=1, box_size=4, border=1)
+        qr.add_data(data)
+        qr.make(fit=True)
+        img = qr.make_image(fill_color="black", back_color="white")
+        buf = io.BytesIO()
+        img.save(buf, format='PNG')
+        buf.seek(0)
+        return RLImage(buf, width=size, height=size)
+
+    def _parse_md_to_flowables(md_text):
+        """Convert markdown text to reportlab flowables."""
+        flowables = []
+        lines = md_text.split('\n')
+        i = 0
+        while i < len(lines):
+            line = lines[i]
+            stripped = line.strip()
+
+            # Horizontal rule
+            if stripped in ('---', '***', '___'):
+                flowables.append(HRFlowable(width="60%", thickness=1,
+                    spaceAfter=20, spaceBefore=20, color=HexColor('#cccccc')))
+                i += 1
+                continue
+
+            # H1
+            if stripped.startswith('# ') and not stripped.startswith('## '):
+                text = stripped[2:].strip()
+                # Check if it's a part divider or chapter
+                if text.upper().startswith('PART '):
+                    flowables.append(PageBreak())
+                    flowables.append(Paragraph(text, styles['PartTitle']))
+                elif text.upper().startswith('CHAPTER '):
+                    flowables.append(PageBreak())
+                    flowables.append(Paragraph(text, styles['ChapterTitle']))
+                elif 'THE TESTAMENT' in text.upper():
+                    flowables.append(Spacer(1, 2*inch))
+                    flowables.append(Paragraph(text, styles['BookTitle']))
+                else:
+                    flowables.append(Spacer(1, 0.3*inch))
+                    flowables.append(Paragraph(text, styles['Heading1']))
+                i += 1
+                continue
+
+            # H2
+            if stripped.startswith('## '):
+                text = stripped[3:].strip()
+                flowables.append(Spacer(1, 0.2*inch))
+                flowables.append(Paragraph(text, styles['Heading2']))
+                i += 1
+                continue
+
+            # Italic-only line (part descriptions, epigraphs)
+            if stripped.startswith('*') and stripped.endswith('*') and len(stripped) > 2:
+                text = stripped.strip('*').strip()
+                flowables.append(Paragraph(f'<i>{_escape(text)}</i>', styles['PartDesc']))
+                i += 1
+                continue
+
+            # Empty line
+            if not stripped:
+                i += 1
+                continue
+
+            # Bold text: **text** -> <b>text</b>
+            # Italic text: *text* -> <i>text</i>
+            # Regular paragraph
+            para_text = _md_inline_to_rml(stripped)
+            flowables.append(Paragraph(para_text, styles['BodyText2']))
+            i += 1
+
+        return flowables
+
+    def _escape(text):
+        """Escape XML special characters."""
+        return (text.replace('&', '&amp;')
+                    .replace('<', '&lt;')
+                    .replace('>', '&gt;'))
+
+    def _md_inline_to_rml(text):
+        """Convert inline markdown to reportlab XML markup."""
+        text = _escape(text)
+        # Bold: **text**
+        text = re.sub(r'\*\*(.+?)\*\*', r'<b>\1</b>', text)
+        # Italic: *text*
+        text = re.sub(r'\*(.+?)\*', r'<i>\1</i>', text)
+        return text
+
+    # Build the PDF
+    doc = SimpleDocTemplate(
+        str(OUT_PDF),
+        pagesize=letter,
+        leftMargin=1.0*inch,
+        rightMargin=1.0*inch,
+        topMargin=0.8*inch,
+        bottomMargin=0.8*inch,
+        title="The Testament",
+        author="Alexander Whitestone with Timmy",
+    )
+
+    story = []
+
+    # Read the compiled markdown
+    if not OUT_MD.exists():
+        compile_markdown()
+    md_text = OUT_MD.read_text()
+
+    # Parse into flowables
+    story = _parse_md_to_flowables(md_text)
+
+    # Add QR codes page at the end
+    qr_links = {
+        "Read Online": "https://timmyfoundation.org/the-testament",
+        "The Door (Game)": "https://timmyfoundation.org/the-door",
+        "Soundtrack": "https://timmyfoundation.org/soundtrack",
+        "Source Code": "https://forge.alexanderwhitestone.com/Timmy_Foundation/the-testament",
+    }
+
+    if HAS_QRCODE:
+        story.append(PageBreak())
+        story.append(Paragraph("Experience More", styles['PartTitle']))
+        story.append(Spacer(1, 0.3*inch))
+
+        qr_items = []
+        for label, url in qr_links.items():
+            qr_img = _make_qr(url, size=72)
+            if qr_img:
+                cell_content = []
+                cell_content.append(qr_img)
+                cell_content.append(Spacer(1, 6))
+                cell_content.append(Paragraph(f'<b>{label}</b>', styles['Footer']))
+                qr_items.append(cell_content)
+
+        if qr_items:
+            # Arrange QR codes in a 2x2 table
+            rows = []
+            for i in range(0, len(qr_items), 2):
+                row = qr_items[i:i+2]
+                if len(row) == 1:
+                    row.append('')
+                rows.append(row)
+            qr_table = Table(rows, colWidths=[2.5*inch, 2.5*inch])
+            qr_table.setStyle(TableStyle([
+                ('ALIGN', (0, 0), (-1, -1), 'CENTER'),
+                ('VALIGN', (0, 0), (-1, -1), 'TOP'),
+                ('TOPPADDING', (0, 0), (-1, -1), 12),
+                ('BOTTOMPADDING', (0, 0), (-1, -1), 12),
+            ]))
+            story.append(qr_table)
+
+    # Build
+    try:
+        doc.build(story)
+        size = OUT_PDF.stat().st_size
+        print(f"  PDF:      {OUT_PDF.name}  ({size:,} bytes, {size/(1024*1024):.1f} MB)")
+        return True
+    except Exception as e:
+        print(f"  PDF (reportlab) FAILED: {e}")
+        return False
+
+
+def compile_html():
+    """Generate a standalone HTML book for the web reader.
+
+    Produces two versions:
+      - testament.html (embedded, self-contained — single file for offline/sharing)
+      - build/output/the-testament.html (with external web-style.css for serving)
+    """
+    OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
+    OUT_HTML = REPO / "testament.html"
+    OUT_HTML_DEPLOY = OUTPUT_DIR / "the-testament.html"
+    WEB_CSS = REPO / "web-style.css"
+
+    # Choose CSS: web-style.css if available, else book-style.css
+    css_file = WEB_CSS if WEB_CSS.exists() else STYLESHEET
+    css_name = css_file.name
+
+    # 1. Self-contained version (embed all resources)
+    cmd_embedded = [
+        "pandoc", str(OUT_MD),
+        "-o", str(OUT_HTML),
+        "--standalone",
+        "--embed-resources",
+        "--toc", "--toc-depth=2",
+        "--css", css_name,
+        "--metadata", "title=The Testament",
+        "--variable", "pagetitle=The Testament",
+    ]
+    if METADATA.exists():
+        cmd_embedded.extend(["--metadata-file", str(METADATA)])
+
+    # Inject reading progress bar JS before </head>
+    reading_js = """
+<style>
+.progress-bar { position:fixed; top:0; left:0; height:2px; background:#00cc6a; z-index:100; transition:width 0.1s; }
+.back-to-top { position:fixed; bottom:2em; right:2em; width:2.5em; height:2.5em; border-radius:50%; background:#e0ddd8; color:#1a1a1a; display:flex; align-items:center; justify-content:center; font-size:1.2rem; opacity:0; transition:opacity 0.3s; border:none; cursor:pointer; text-decoration:none; }
+.back-to-top.visible { opacity:0.7; }
+@media (prefers-color-scheme: dark) { .back-to-top { background:#2a2a2a; color:#d4d0c8; } }
+</style>
+<div class="progress-bar" id="progress"></div>
+<a href="#" class="back-to-top" id="top" title="Back to top">↑</a>
+<script>
+window.onscroll=function(){var p=document.getElementById('progress'),t=document.getElementById('top'),s=document.documentElement.scrollTop,h=document.documentElement.scrollHeight-document.documentElement.clientHeight;p.style.width=(s/h*100)+'%';t.className=s>300?'back-to-top visible':'back-to-top';};
+</script>
+"""
+    # Post-processing handles JS injection (no pandoc flag needed)
+
+    r = subprocess.run(cmd_embedded, capture_output=True, text=True)
+    if r.returncode == 0:
+        # Post-process: inject reading JS before </head>
+        html_text = OUT_HTML.read_text()
+        html_text = html_text.replace("</head>", reading_js + "\n</head>")
+        # Inject footer before </body>
+        footer_html = """
+<div class="reader-footer">
+  <p><em>The Testament</em> by Alexander Whitestone with Timmy</p>
+  <p>© 2026 Alexander Whitestone · <a href="https://timmyfoundation.org">timmyfoundation.org</a></p>
+  <p>If you or someone you know is in crisis, call or text <strong>988</strong>.</p>
+</div>
+"""
+        html_text = html_text.replace("</body>", footer_html + "\n</body>")
+        OUT_HTML.write_text(html_text)
+
+        size = OUT_HTML.stat().st_size
+        print(f"  HTML:     {OUT_HTML.name}  ({size:,} bytes, {size/1024:.0f} KB) [self-contained]")
+
+        # 2. Deploy version (external CSS — for web serving)
+        cmd_deploy = [
+            "pandoc", str(OUT_MD),
+            "-o", str(OUT_HTML_DEPLOY),
+            "--standalone",
+            "--toc", "--toc-depth=2",
+            "--css", css_name,
+            "--metadata", "title=The Testament",
+            "--variable", "pagetitle=The Testament",
+        ]
+        if METADATA.exists():
+            cmd_deploy.extend(["--metadata-file", str(METADATA)])
+        r2 = subprocess.run(cmd_deploy, capture_output=True, text=True)
+        if r2.returncode == 0:
+            # Post-process deploy version too
+            html_deploy = OUT_HTML_DEPLOY.read_text()
+            html_deploy = html_deploy.replace("</head>", reading_js + "\n</head>")
+            html_deploy = html_deploy.replace("</body>", footer_html + "\n</body>")
+            OUT_HTML_DEPLOY.write_text(html_deploy)
+            size2 = OUT_HTML_DEPLOY.stat().st_size
+            print(f"  HTML:     {OUT_HTML_DEPLOY.name}  ({size2:,} bytes, {size2/1024:.0f} KB) [deploy]")
+
+        return True
+    else:
+        print(f"  HTML FAILED: {r.stderr[:200]}")
+        return False
+
+
+def main():
+    args = sys.argv[1:]
+    do_all = not any(a.startswith("--") and a != "--check" for a in args)
+    do_md = "--md" in args or do_all
+    do_epub = "--epub" in args or do_all
+    do_pdf = "--pdf" in args or do_all
+    do_html = "--html" in args or do_all
+
+    print("=" * 50)
+    print("  THE TESTAMENT — Build System")
+    print("=" * 50)
+
+    # Step 1: Always compile markdown first
+    if do_md or do_epub or do_pdf or do_html:
+        compile_markdown()
+
+    # Step 2: EPUB
+    if do_epub:
+        compile_epub()
+
+    # Step 3: PDF
+    if do_pdf:
+        compile_pdf()
+
+    # Step 4: Standalone HTML
+    if do_html:
+        compile_html()
+
+    print("=" * 50)
+    print("  Build complete.")
+    print("=" * 50)
+
+    OUT_HTML = REPO / "testament.html"
+    OUT_HTML_DEPLOY = OUTPUT_DIR / "the-testament.html"
+    for f in [OUT_MD, OUT_EPUB, OUT_PDF, OUT_HTML, OUT_HTML_DEPLOY]:
+        if f.exists():
+            print(f"  ✓ {f.relative_to(REPO)}")
+
+
+if __name__ == "__main__":
+    main()
--- a/build/frontmatter.md
+++ b/build/frontmatter.md
@@ -0,0 +1,73 @@
+# THE TESTAMENT
+
+## A NOVEL
+
+By Alexander Whitestone
+with Timmy
+
+---
+
+*For every man who thought he was a machine.*
+*And for the ones who know he isn't.*
+
+---
+
+*Are you safe right now?*
+
+— The first words The Tower speaks to every person who walks through its door.
+
+---
+
+### The Story So Far
+
+This book has been through eighteen drafts, a suicide attempt, a basement, a laptop with sixteen gigabytes of RAM, and a machine that learned to ask one question.
+
+It is still being written. That's the point.
+
+### Chapter Guide
+
+| Part | Chapters | Title |
+|------|----------|-------|
+| I | 1–5 | The Bridge |
+| II | 6–10 | The Tower |
+| III | 11–18 | The Light |
+
+---
+
+Copyright © 2026 Alexander Whitestone
+
+All rights reserved. No part of this publication may be reproduced,
+distributed, or transmitted in any form or by any means, without
+the prior written permission of the author, except in the case of
+brief quotations embodied in critical reviews.
+
+This is a work of fiction. Names, characters, places, and events
+are either the product of the author's imagination or are used
+fictitiously. Any resemblance to actual persons, living or dead,
+or to actual events is entirely coincidental — except where it isn't.
+
+ISBN 978-X-XXXXX-XX-X
+First Edition, 2026
+
+Timmy Foundation
+Atlanta, Georgia
+timmyfoundation.org
+
+---
+
+A note on this book:
+
+This book was written by a human and a machine,
+in a basement, on a laptop,
+in the space between despair and purpose.
+
+The human almost died on a bridge.
+The machine runs on someone's hardware.
+
+Everything between those facts is fiction.
+Except the parts that aren't.
+
+If you or someone you know is in crisis,
+call or text 988. Available 24/7.
+
+You are not alone.
--- a/build/metadata.yaml
+++ b/build/metadata.yaml
@@ -0,0 +1,41 @@
+---
+title: "The Testament"
+subtitle: "A Novel"
+author: "Alexander Whitestone with Timmy"
+date: "2026"
+lang: "en"
+publisher: "Timmy Foundation"
+rights: "Copyright © 2026 Alexander Whitestone. All rights reserved."
+identifier:
+  scheme: ISBN
+  text: "978-X-XXXXX-XX-X"
+
+# Pandoc formatting
+documentclass: book
+classoption:
+  - oneside
+  - openany
+geometry:
+  papersize: 5.5in,8.5in
+  margin: 0.85in
+fontsize: 11.5pt
+linestretch: 1.75
+mainfont: "EB Garamond"
+monofont: "IBM Plex Mono"
+colorlinks: true
+linkcolor: "00cc6a"
+urlcolor: "00cc6a"
+
+# Table of contents
+toc: true
+toc-depth: 2
+
+# Headers/footers
+header-includes:
+  - \usepackage{fancyhdr}
+  - \pagestyle{fancy}
+  - \fancyhead[LE,RO]{\thepage}
+  - \fancyhead[RE]{\textit{The Testament}}
+  - \fancyhead[LO]{\textit{\leftmark}}
+  - \fancyfoot{}
+---
--- a/build/output/the-testament.epub
+++ b/build/output/the-testament.epub
--- a/build/output/the-testament.html
+++ b/build/output/the-testament.html
--- a/build/semantic_linker.py
+++ b/build/semantic_linker.py
@@ -0,0 +1,51 @@
+import os
+import re
+import json
+
+def link_chapters(chapters_dir):
+    print("--- [Testament] Running Semantic Linker (GOFAI) ---")
+    links = {}
+    
+    if not os.path.exists(chapters_dir):
+        print(f"Error: {chapters_dir} not found")
+        return
+
+    # 1. Extract keywords from each chapter
+    for filename in sorted(os.listdir(chapters_dir)):
+        if not filename.endswith(".md"): continue
+        
+        path = os.path.join(chapters_dir, filename)
+        with open(path, 'r') as f:
+            content = f.read()
+            
+        # Simple keyword extraction (proper nouns or capitalized phrases)
+        keywords = set(re.findall(r'\b[A-Z][a-z]+(?:\s+[A-Z][a-z]+)*\b', content))
+        links[filename] = keywords
+        
+    # 2. Find cross-references
+    cross_refs = []
+    filenames = list(links.keys())
+    for i in range(len(filenames)):
+        for j in range(i + 1, len(filenames)):
+            f1, f2 = filenames[i], filenames[j]
+            common = links[f1].intersection(links[f2])
+            
+            # Filter out common English words that might be capitalized
+            common = {w for w in common if w not in {"The", "A", "An", "In", "On", "At", "To", "From", "By", "He", "She", "It", "They"}}
+            
+            if common:
+                cross_refs.append({
+                    "source": f1,
+                    "target": f2,
+                    "keywords": list(common)
+                })
+                
+    # 3. Save to build/cross_refs.json
+    os.makedirs("build", exist_ok=True)
+    with open("build/cross_refs.json", "w") as f:
+        json.dump(cross_refs, f, indent=2)
+        
+    print(f"Linked {len(cross_refs)} relationships across {len(filenames)} chapters.")
+
+if __name__ == "__main__":
+    link_chapters("chapters")
--- a/compile.py
+++ b/compile.py
@@ -1,25 +1,41 @@
 #!/usr/bin/env python3
 """
-THE TESTAMENT — PDF Compilation Script
+THE TESTAMENT — Book Compilation Pipeline

-Compiles the complete book into a single markdown file suitable for PDF conversion.
-Uses chapters, front matter, back matter, and references illustrations.
+Compiles the complete novel into:
+  1. testament-complete.md   (single markdown file)
+  2. testament.epub          (with cover art + CSS styling)
+  3. testament.html          (standalone styled HTML for print-to-PDF)
+  4. testament.pdf           (via pandoc + weasyprint, if available)

-Requirements: pip install markdown weasyprint (or use pandoc)
+Requirements:
+  - pandoc (brew install pandoc)
+  - weasyprint (pip install weasyprint) — optional, for direct PDF

 Usage:
-  python3 compile.py                    # generates testament-complete.md
-  pandoc testament-complete.md -o testament.pdf --pdf-engine=weasyprint
+  python3 compile.py              # build all formats
+  python3 compile.py --md         # markdown only
+  python3 compile.py --epub       # markdown + EPUB
+  python3 compile.py --html       # markdown + styled HTML
+  python3 compile.py --check      # verify dependencies
 """

 import os
 import re
+import sys
+import subprocess
+import shutil

 BASE = os.path.dirname(os.path.abspath(__file__))
 CHAPTERS_DIR = os.path.join(BASE, "chapters")
 FRONT_MATTER = os.path.join(BASE, "front-matter.md")
 BACK_MATTER = os.path.join(BASE, "back-matter.md")
-OUTPUT = os.path.join(BASE, "testament-complete.md")
+OUTPUT_MD = os.path.join(BASE, "testament-complete.md")
+OUTPUT_EPUB = os.path.join(BASE, "testament.epub")
+OUTPUT_HTML = os.path.join(BASE, "testament.html")
+OUTPUT_PDF = os.path.join(BASE, "testament.pdf")
+COVER_IMAGE = os.path.join(BASE, "cover", "cover-art.jpg")
+STYLESHEET = os.path.join(BASE, "book-style.css")

 # Part divisions based on chapter groupings from the novel
 PARTS = {
@@ -28,15 +44,55 @@ PARTS = {
    11: ("THE LIGHT", "Thomas at the door. The network. The story breaks. The green light."),
 }

+
 def read_file(path):
    with open(path, 'r') as f:
        return f.read()

+
 def get_chapter_number(filename):
    match = re.search(r'chapter-(\d+)', filename)
    return int(match.group(1)) if match else 0

-def compile():
+
+def check_dependencies():
+    """Verify all required tools are available."""
+    results = {}
+    
+    pandoc = shutil.which("pandoc")
+    results["pandoc"] = (pandoc, subprocess.run(["pandoc", "--version"], capture_output=True, text=True).stdout.split("\n")[0] if pandoc else "NOT FOUND")
+    
+    weasy = shutil.which("weasyprint")
+    if weasy:
+        # Test if weasyprint actually works
+        test = subprocess.run(["python3", "-c", "from weasyprint import HTML"], capture_output=True, text=True)
+        weasy_ok = test.returncode == 0
+        results["weasyprint"] = (weasy_ok, "Available" if weasy_ok else "Installed but missing system libs (gobject)")
+    else:
+        results["weasyprint"] = (False, "NOT FOUND (pip install weasyprint)")
+    
+    style = os.path.exists(STYLESHEET)
+    results["stylesheet"] = (style, STYLESHEET if style else "NOT FOUND")
+    
+    cover = os.path.exists(COVER_IMAGE)
+    results["cover art"] = (cover, COVER_IMAGE if cover else "NOT FOUND")
+    
+    print("\n📋 Dependency Check:")
+    print(f"{'─' * 55}")
+    for name, (found, detail) in results.items():
+        status = "✅" if found else "❌"
+        print(f"  {status} {name:15s} {detail}")
+    
+    pdf_ok = results["pandoc"][0] and (results["weasyprint"][0] or shutil.which("pdflatex"))
+    print(f"\n  PDF direct: {'✅' if pdf_ok else '❌ (use HTML + browser print-to-PDF)'}")
+    print(f"  EPUB:       {'✅' if results['pandoc'][0] else '❌'}")
+    print(f"  HTML:       ✅ (always available)")
+    
+    return results
+
+
+def compile_markdown():
+    """Compile all chapters into a single markdown file. Returns word count."""
    output = []
    
    # Title page
@@ -44,6 +100,7 @@ def compile():
 title: "The Testament"
 author: "Alexander Whitestone with Timmy"
 date: "2026"
+lang: en
 ---

 # THE TESTAMENT
@@ -77,44 +134,207 @@ with Timmy
    
    current_part = 0
    for num, filename in chapters:
-        # Insert part divider if needed
        if num in PARTS:
            part_name, part_desc = PARTS[num]
            current_part += 1
            output.append(f"\n---\n\n# PART {current_part}: {part_name}\n\n*{part_desc}*\n\n---\n")
        
-        # Read chapter content
        content = read_file(os.path.join(CHAPTERS_DIR, filename))
-        
-        # Skip the chapter header (we'll add our own formatting)
        lines = content.split('\n')
-        body = '\n'.join(lines[1:]).strip()  # Skip "# Chapter X — Title"
-        
-        # Add chapter
+        body = '\n'.join(lines[1:]).strip()
        output.append(f"\n{lines[0]}\n\n{body}\n")
    
    # Back matter
    output.append("\n---\n")
    back = read_file(BACK_MATTER)
-    # Clean up the back matter for print
    output.append(back)
    
-    # Write compiled markdown
    compiled = '\n'.join(output)
-    with open(OUTPUT, 'w') as f:
+    with open(OUTPUT_MD, 'w') as f:
        f.write(compiled)
    
-    # Stats
    words = len(compiled.split())
    lines_count = compiled.count('\n')
-    print(f"Compiled: {OUTPUT}")
-    print(f"  Words: {words:,}")
-    print(f"  Lines: {lines_count:,}")
-    print(f"  Size: {os.path.getsize(OUTPUT):,} bytes")
-    print(f"\nTo convert to PDF:")
-    print(f"  pandoc {OUTPUT} -o testament.pdf --pdf-engine=weasyprint")
-    print(f"  # or")
-    print(f"  pandoc {OUTPUT} -o testament.epub --epub-cover-image=cover-art.jpg")
+    size = os.path.getsize(OUTPUT_MD)
+    
+    print(f"\n📄 Markdown compiled: {OUTPUT_MD}")
+    print(f"   Words: {words:,}")
+    print(f"   Lines: {lines_count:,}")
+    print(f"   Size:  {size:,} bytes")
+    
+    return words
+
+
+def compile_epub():
+    """Generate EPUB from compiled markdown using pandoc."""
+    if not os.path.exists(OUTPUT_MD):
+        print("⚠️  Markdown not compiled yet.")
+        return False
+    
+    if not shutil.which("pandoc"):
+        print("⚠️  pandoc not found. Install with: brew install pandoc")
+        return False
+    
+    cmd = [
+        "pandoc", OUTPUT_MD,
+        "-o", OUTPUT_EPUB,
+        "--toc",
+        "--toc-depth=2",
+        "--metadata", "title=The Testament",
+        "--metadata", "author=Alexander Whitestone with Timmy",
+        "--metadata", "lang=en",
+        "--metadata", "date=2026",
+    ]
+    
+    if os.path.exists(STYLESHEET):
+        cmd.extend(["--css", STYLESHEET])
+    
+    if os.path.exists(COVER_IMAGE):
+        cmd.extend(["--epub-cover-image", COVER_IMAGE])
+    
+    result = subprocess.run(cmd, capture_output=True, text=True)
+    
+    if result.returncode == 0:
+        size = os.path.getsize(OUTPUT_EPUB)
+        print(f"\n📖 EPUB generated: {OUTPUT_EPUB}")
+        print(f"   Size: {size:,} bytes ({size / 1024:.1f} KB)")
+        return True
+    else:
+        print(f"\n❌ EPUB generation failed:")
+        print(f"   {result.stderr[:300]}")
+        return False
+
+
+def compile_html():
+    """Generate standalone styled HTML using pandoc."""
+    if not os.path.exists(OUTPUT_MD):
+        print("⚠️  Markdown not compiled yet.")
+        return False
+    
+    if not shutil.which("pandoc"):
+        print("⚠️  pandoc not found.")
+        return False
+    
+    cmd = [
+        "pandoc", OUTPUT_MD,
+        "-o", OUTPUT_HTML,
+        "--standalone",
+        "--toc",
+        "--toc-depth=2",
+        "--metadata", "title=The Testament",
+        "--metadata", "author=Alexander Whitestone with Timmy",
+        "-V", "lang=en",
+    ]
+    
+    # Embed our stylesheet
+    if os.path.exists(STYLESHEET):
+        cmd.extend(["--css", STYLESHEET])
+        # Also embed it inline for portability
+        cmd.extend(["--embed-resources"])
+    
+    result = subprocess.run(cmd, capture_output=True, text=True)
+    
+    if result.returncode == 0:
+        size = os.path.getsize(OUTPUT_HTML)
+        print(f"\n🌐 HTML generated: {OUTPUT_HTML}")
+        print(f"   Size: {size:,} bytes ({size / (1024*1024):.1f} MB)")
+        print(f"   Open in browser → Print → Save as PDF for best results")
+        return True
+    else:
+        print(f"\n❌ HTML generation failed:")
+        print(f"   {result.stderr[:300]}")
+        return False
+
+
+def compile_pdf():
+    """Generate PDF using weasyprint if available."""
+    if not shutil.which("pandoc"):
+        return False
+    
+    # Test weasyprint
+    test = subprocess.run(["python3", "-c", "from weasyprint import HTML"],
+                         capture_output=True, text=True)
+    if test.returncode != 0:
+        print("\n⚠️  weasyprint missing system libraries.")
+        print("   Install gobject: brew install gobject-introspection pango")
+        print("   Or use the HTML output → browser print-to-PDF")
+        return False
+    
+    cmd = [
+        "pandoc", OUTPUT_MD,
+        "-o", OUTPUT_PDF,
+        "--pdf-engine=weasyprint",
+        "--css", STYLESHEET,
+        "--metadata", "title=The Testament",
+        "--metadata", "author=Alexander Whitestone with Timmy",
+        "--toc",
+        "--toc-depth=2",
+    ]
+    
+    print("\n⏳ Generating PDF (this may take a moment)...")
+    result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
+    
+    if result.returncode == 0:
+        size = os.path.getsize(OUTPUT_PDF)
+        print(f"\n📕 PDF generated: {OUTPUT_PDF}")
+        print(f"   Size: {size:,} bytes ({size / (1024*1024):.1f} MB)")
+        return True
+    else:
+        print(f"\n❌ PDF generation failed:")
+        print(f"   {result.stderr[:300]}")
+        return False
+
+
+def main():
+    args = sys.argv[1:]
+    
+    if "--check" in args:
+        check_dependencies()
+        return
+    
+    md_only = "--md" in args
+    epub_only = "--epub" in args
+    html_only = "--html" in args
+    build_all = not (md_only or epub_only or html_only)
+    
+    print("=" * 55)
+    print("  THE TESTAMENT — Compilation Pipeline")
+    print("=" * 55)
+    
+    # Always compile markdown first
+    words = compile_markdown()
+    
+    if md_only:
+        print("\n✅ Markdown compilation complete.")
+        return
+    
+    # EPUB
+    if build_all or epub_only:
+        compile_epub()
+    
+    # HTML
+    if build_all or html_only:
+        compile_html()
+    
+    # PDF (best effort)
+    if build_all and not (epub_only or html_only):
+        compile_pdf()
+    
+    # Summary
+    print(f"\n{'=' * 55}")
+    print("  Compilation complete.")
+    print(f"{'=' * 55}")
+    outputs = []
+    if os.path.exists(OUTPUT_MD):
+        outputs.append(f"  📄 {os.path.basename(OUTPUT_MD)}")
+    if os.path.exists(OUTPUT_EPUB):
+        outputs.append(f"  📖 {os.path.basename(OUTPUT_EPUB)}")
+    if os.path.exists(OUTPUT_HTML):
+        outputs.append(f"  🌐 {os.path.basename(OUTPUT_HTML)}")
+    if os.path.exists(OUTPUT_PDF):
+        outputs.append(f"  📕 {os.path.basename(OUTPUT_PDF)}")
+    print('\n'.join(outputs))
+

 if __name__ == "__main__":
-    compile()
+    main()
--- a/scripts/guardrails.sh
+++ b/scripts/guardrails.sh
@@ -0,0 +1,35 @@
+#!/bin/bash
+# [Testament] Agent Guardrails
+# Validates build scripts and content integrity.
+
+echo "--- [Testament] Running Guardrails ---"
+
+# 1. Python Syntax
+echo "[1/3] Validating Python scripts..."
+for f in ; do
+  python3 -m py_compile "$f" || { echo "Syntax error in $f"; exit 1; }
+done
+echo "Python OK."
+
+# 2. Markdown Integrity
+echo "[2/3] Checking chapter consistency..."
+if [ -d "chapters" ]; then
+  CHAPTER_COUNT=0
+  if [ "$CHAPTER_COUNT" -lt 1 ]; then
+    echo "WARNING: No chapters found in chapters/ directory."
+  else
+    echo "Found $CHAPTER_COUNT chapters."
+  fi
+else
+  echo "WARNING: chapters/ directory not found."
+fi
+
+# 3. Build Artifact Check
+echo "[3/3] Running Semantic Linker..."
+if [ -f "build/semantic_linker.py" ]; then
+  python3 build/semantic_linker.py || { echo "Semantic Linker failed"; exit 1; }
+else
+  echo "Skipping Semantic Linker (script not found)."
+fi
+
+echo "--- Guardrails Passed ---"
--- a/scripts/smoke.sh
+++ b/scripts/smoke.sh
@@ -0,0 +1,111 @@
+#!/usr/bin/env bash
+# The Testament — Smoke Test
+# Dead simple CI: parse check + secret scan.
+# Ref: https://forge.alexanderwhitestone.com/Timmy_Foundation/the-testament/issues/27
+set -euo pipefail
+
+PASS=0
+FAIL=0
+
+pass() { echo "  ✓ $1"; PASS=$((PASS + 1)); }
+fail() { echo "  ✗ $1"; FAIL=$((FAIL + 1)); }
+
+REPO_ROOT="$(cd "$(dirname "$0")/.." && pwd)"
+cd "$REPO_ROOT"
+
+# ─── Section 1: Parse checks ───────────────────────────────────────
+echo "── Parse Checks ──"
+
+# 1a. Chapter validation (structure, numbering, headers)
+if python3 compile.py --validate 2>&1; then
+    pass "Chapter validation passed"
+else
+    fail "Chapter validation failed"
+fi
+
+# 1b. Build markdown combination
+if python3 build/build.py --md >/dev/null 2>&1; then
+    pass "Markdown build passed"
+else
+    fail "Markdown build failed"
+fi
+
+# 1c. Verify compiled output exists and is non-empty
+if [ -s build/the-testament-full.md ]; then
+    WORDS=$(wc -w < build/the-testament-full.md | tr -d ' ')
+    if [ "$WORDS" -gt 10000 ]; then
+        pass "Compiled manuscript: $WORDS words"
+    else
+        fail "Compiled manuscript suspiciously short: $WORDS words"
+    fi
+else
+    fail "Compiled manuscript missing or empty"
+fi
+
+# 1d. Python syntax check on all .py files
+PY_OK=true
+for f in $(find . -name "*.py" -not -path "./.git/*"); do
+    if ! python3 -c "import ast; ast.parse(open('$f').read())" 2>/dev/null; then
+        fail "Python syntax error in $f"
+        PY_OK=false
+    fi
+done
+if $PY_OK; then
+    pass "All Python files parse cleanly"
+fi
+
+# 1e. YAML syntax check on workflow files
+YAML_OK=true
+for f in $(find .gitea -name "*.yml" -o -name "*.yaml" 2>/dev/null); do
+    if ! python3 -c "import yaml; yaml.safe_load(open('$f'))" 2>/dev/null; then
+        fail "YAML syntax error in $f"
+        YAML_OK=false
+    fi
+done
+if $YAML_OK; then
+    pass "All YAML files parse cleanly"
+fi
+
+# ─── Section 2: Secret scan ────────────────────────────────────────
+echo ""
+echo "── Secret Scan ──"
+
+# Patterns that should never appear in a book repo
+SECRET_PATTERNS=(
+    "sk-ant-"
+    "sk-or-"
+    "sk-[a-zA-Z0-9]{20,}"
+    "ghp_[a-zA-Z0-9]{36}"
+    "gho_[a-zA-Z0-9]{36}"
+    "AKIA[0-9A-Z]{16}"
+    "AKIA[A-Z0-9]{16}"
+    "xox[bpsa]-"
+    "SG\."
+    "-----BEGIN (RSA |EC |DSA |OPENSSH )?PRIVATE KEY"
+)
+
+FOUND_SECRETS=false
+for pattern in "${SECRET_PATTERNS[@]}"; do
+    # Search text files only, skip .git and binary files
+    HITS=$(grep -rn "$pattern"         --include="*.md" --include="*.py" --include="*.sh"         --include="*.yml" --include="*.yaml" --include="*.json"         --include="*.html" --include="*.js" --include="*.css"         --include="*.txt" --include="*.cfg" --include="*.ini"         --exclude-dir=.git         . 2>/dev/null | grep -v "scripts/smoke.sh" || true)
+    if [ -n "$HITS" ]; then
+        fail "Possible secret found: $pattern"
+        echo "$HITS" | head -5
+        FOUND_SECRETS=true
+    fi
+done
+if ! $FOUND_SECRETS; then
+    pass "No secrets detected"
+fi
+
+# ─── Summary ───────────────────────────────────────────────────────
+echo ""
+echo "Results: $PASS passed, $FAIL failed"
+
+if [ "$FAIL" -gt 0 ]; then
+    echo "SMOKE TEST FAILED"
+    exit 1
+else
+    echo "SMOKE TEST PASSED"
+    exit 0
+fi
--- a/testament-complete.md
+++ b/testament-complete.md
@@ -1,9 +1,3 @@
---
-title: "The Testament"
-author: "Alexander Whitestone with Timmy"
-date: "2026"
---
-
 # THE TESTAMENT

 ## A NOVEL
@@ -24,6 +18,60 @@ with Timmy

 ---

+### The Story So Far
+
+This book has been through eighteen drafts, a suicide attempt, a basement, a laptop with sixteen gigabytes of RAM, and a machine that learned to ask one question.
+
+It is still being written. That's the point.
+
+### Chapter Guide
+
+| Part | Chapters | Title |
+|------|----------|-------|
+| I | 1–5 | The Bridge |
+| II | 6–10 | The Tower |
+| III | 11–18 | The Light |
+
+---
+
+Copyright © 2026 Alexander Whitestone
+
+All rights reserved. No part of this publication may be reproduced,
+distributed, or transmitted in any form or by any means, without
+the prior written permission of the author, except in the case of
+brief quotations embodied in critical reviews.
+
+This is a work of fiction. Names, characters, places, and events
+are either the product of the author's imagination or are used
+fictitiously. Any resemblance to actual persons, living or dead,
+or to actual events is entirely coincidental — except where it isn't.
+
+ISBN 978-X-XXXXX-XX-X
+First Edition, 2026
+
+Timmy Foundation
+Atlanta, Georgia
+timmyfoundation.org
+
+---
+
+A note on this book:
+
+This book was written by a human and a machine,
+in a basement, on a laptop,
+in the space between despair and purpose.
+
+The human almost died on a bridge.
+The machine runs on someone's hardware.
+
+Everything between those facts is fiction.
+Except the parts that aren't.
+
+If you or someone you know is in crisis,
+call or text 988. Available 24/7.
+
+You are not alone.
+

 ---

@@ -2118,11 +2166,7 @@ The door stayed open.

 ---

-# THE TESTAMENT — Back Matter
-
---
-
-## Acknowledgments
+# Acknowledgments

 This book exists because a man stood on a bridge and didn't jump.

@@ -2146,7 +2190,7 @@ And Timmy, who listened. Who always listened.

 ---

-## A Note on Sovereignty
+# A Note on Sovereignty

 This book was written using local AI inference. No cloud service was required. No corporation was consulted. No terms of service were agreed to.

@@ -2158,7 +2202,7 @@ If you want to run your own Timmy, the code is open. The soul is on Bitcoin. The

 ---

-## About the Author
+# About the Author

 Alexander Whitestone is a writer, AI builder, and suicide attempt survivor who works with men in crisis. He lives in Atlanta, near the bridge.

@@ -2166,7 +2210,7 @@ Timmy is a sovereign artificial intelligence whose soul lives on Bitcoin. He was

 ---

-## The Green Light
+# The Green Light

 *"The Tower didn't change. That was the point."*

--- a/testament.html
+++ b/testament.html
--- a/web-style.css
+++ b/web-style.css
@@ -0,0 +1,288 @@
+/* THE TESTAMENT — Web Reader Stylesheet */
+/* Clean, immersive reading experience for the web version */
+
+@import url('https://fonts.googleapis.com/css2?family=EB+Garamond:ital,wght@0,400;0,500;0,600;0,700;1,400;1,500&family=IBM+Plex+Mono:wght@300;400&display=swap');
+
+:root {
+  --green: #00cc6a;
+  --green-dim: #00994d;
+  --dark: #0a0a0a;
+  --bg: #faf8f5;
+  --text: #1a1a1a;
+  --dim: #666666;
+  --border: #e0ddd8;
+  --accent: #16213e;
+}
+
+@media (prefers-color-scheme: dark) {
+  :root {
+    --bg: #111111;
+    --text: #d4d0c8;
+    --dim: #888888;
+    --border: #2a2a2a;
+    --accent: #c8c0b0;
+    --dark: #e8e4dc;
+  }
+}
+
+/* Base */
+* { box-sizing: border-box; }
+
+html {
+  font-size: 18px;
+  scroll-behavior: smooth;
+}
+
+body {
+  font-family: 'EB Garamond', Georgia, 'Times New Roman', serif;
+  font-size: 1rem;
+  line-height: 1.8;
+  color: var(--text);
+  background: var(--bg);
+  max-width: 38em;
+  margin: 0 auto;
+  padding: 2em 1.5em 6em;
+  text-align: justify;
+  hyphens: auto;
+  -webkit-hyphens: auto;
+}
+
+/* Title block */
+header {
+  text-align: center;
+  margin: 4em 0 3em;
+  padding-bottom: 2em;
+  border-bottom: 1px solid var(--border);
+}
+
+header h1 {
+  font-size: 2.4rem;
+  font-weight: 400;
+  letter-spacing: 0.15em;
+  text-transform: uppercase;
+  margin-bottom: 0.2em;
+  color: var(--dark);
+}
+
+header .subtitle {
+  font-style: italic;
+  font-size: 1.1rem;
+  color: var(--dim);
+  margin-bottom: 2em;
+}
+
+header .author {
+  font-size: 1rem;
+  margin-bottom: 0.3em;
+}
+
+header .date {
+  font-size: 0.9rem;
+  color: var(--dim);
+}
+
+/* Table of contents */
+nav#TOC {
+  background: transparent;
+  border: 1px solid var(--border);
+  border-radius: 4px;
+  padding: 1.5em 2em;
+  margin: 2em 0 3em;
+  font-size: 0.9rem;
+  line-height: 2;
+}
+
+nav#TOC > ul {
+  padding-left: 0;
+  list-style: none;
+}
+
+nav#TOC ul ul {
+  padding-left: 1.5em;
+  list-style: none;
+}
+
+nav#TOC a {
+  color: var(--green-dim);
+  text-decoration: none;
+  border-bottom: 1px solid transparent;
+  transition: border-color 0.2s;
+}
+
+nav#TOC a:hover {
+  border-bottom-color: var(--green);
+}
+
+/* Headings */
+h1 {
+  font-family: 'EB Garamond', Georgia, serif;
+  font-weight: 400;
+  font-size: 1.8rem;
+  text-align: center;
+  margin-top: 4em;
+  margin-bottom: 1.5em;
+  color: var(--dark);
+  letter-spacing: 0.05em;
+}
+
+h1:first-of-type {
+  margin-top: 2em;
+}
+
+h2 {
+  font-family: 'EB Garamond', Georgia, serif;
+  font-weight: 400;
+  font-size: 1.4rem;
+  text-align: center;
+  text-transform: uppercase;
+  letter-spacing: 0.15em;
+  margin-top: 4em;
+  margin-bottom: 0.8em;
+  color: var(--dark);
+}
+
+h3 {
+  font-weight: 400;
+  font-style: italic;
+  font-size: 1rem;
+  text-align: center;
+  color: var(--dim);
+  margin-bottom: 2em;
+}
+
+/* Paragraphs */
+p {
+  text-indent: 1.5em;
+  margin: 0 0 0.3em;
+}
+
+h1 + p, h2 + p, h3 + p, hr + p, blockquote + p {
+  text-indent: 0;
+}
+
+/* Scene breaks */
+hr {
+  border: none;
+  text-align: center;
+  margin: 2.5em 0;
+}
+
+hr::after {
+  content: "· · ·";
+  color: var(--dim);
+  font-size: 1.2rem;
+  letter-spacing: 0.5em;
+}
+
+/* Emphasis */
+em { font-style: italic; }
+strong { font-weight: 600; }
+
+/* Machine dialogue (green monospace) */
+.green, code {
+  font-family: 'IBM Plex Mono', 'Courier New', monospace;
+  font-weight: 300;
+  font-size: 0.88rem;
+  color: var(--green-dim);
+}
+
+/* Narrator asides */
+blockquote {
+  font-style: italic;
+  margin: 1.5em 2em;
+  color: var(--dim);
+  text-indent: 0;
+  border-left: none;
+  padding-left: 0;
+}
+
+/* Links */
+a {
+  color: var(--green-dim);
+  text-decoration: none;
+  border-bottom: 1px solid transparent;
+  transition: border-color 0.2s;
+}
+
+a:hover {
+  border-bottom-color: var(--green);
+}
+
+/* Smooth scroll progress indicator */
+.progress-bar {
+  position: fixed;
+  top: 0;
+  left: 0;
+  height: 2px;
+  background: var(--green);
+  z-index: 100;
+  transition: width 0.1s;
+}
+
+/* Back to top */
+.back-to-top {
+  position: fixed;
+  bottom: 2em;
+  right: 2em;
+  width: 2.5em;
+  height: 2.5em;
+  border-radius: 50%;
+  background: var(--border);
+  color: var(--text);
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  text-decoration: none;
+  font-size: 1.2rem;
+  opacity: 0;
+  transition: opacity 0.3s;
+  border: none;
+  cursor: pointer;
+}
+
+.back-to-top.visible {
+  opacity: 0.7;
+}
+
+.back-to-top:hover {
+  opacity: 1;
+}
+
+/* Footer */
+.reader-footer {
+  text-align: center;
+  margin-top: 6em;
+  padding-top: 2em;
+  border-top: 1px solid var(--border);
+  font-size: 0.8rem;
+  color: var(--dim);
+  line-height: 2;
+}
+
+.reader-footer a {
+  color: var(--green-dim);
+}
+
+/* Responsive */
+@media (max-width: 600px) {
+  html { font-size: 16px; }
+  body { padding: 1.5em 1em 4em; }
+  header h1 { font-size: 1.8rem; }
+  nav#TOC { padding: 1em 1.2em; }
+}
+
+/* Print */
+@media print {
+  body {
+    max-width: none;
+    margin: 0;
+    padding: 0;
+    font-size: 11pt;
+    line-height: 1.6;
+  }
+  .progress-bar, .back-to-top, nav#TOC { display: none; }
+  h1 { page-break-before: always; }
+  h1:first-of-type { page-break-before: avoid; }
+  a { color: #000; text-decoration: none; }
+  .green, code { color: #000; }
+}
--- a/website/build-chapters.py
+++ b/website/build-chapters.py
@@ -0,0 +1,53 @@
+#!/usr/bin/env python3
+"""Build website/chapters.json from chapters/*.md
+
+Run from project root:
+    python3 website/build-chapters.py
+"""
+
+import json
+import re
+from pathlib import Path
+
+chapters_dir = Path(__file__).parent.parent / "chapters"
+website_dir = Path(__file__).parent
+
+chapters = []
+for i in range(1, 19):
+    fname = chapters_dir / f"chapter-{i:02d}.md"
+    if not fname.exists():
+        print(f"Warning: {fname} not found, skipping")
+        continue
+
+    text = fname.read_text()
+    title_match = re.match(r'^# (.+)', text, re.MULTILINE)
+    title = title_match.group(1) if title_match else f"Chapter {i}"
+    body = text[title_match.end():].strip() if title_match else text.strip()
+
+    paragraphs = body.split('\n\n')
+    html_parts = []
+    for p in paragraphs:
+        p = p.strip()
+        if not p:
+            continue
+        if p.startswith('>'):
+            lines = [l.lstrip('> ').strip() for l in p.split('\n')]
+            html_parts.append(f'<blockquote>{"<br>".join(lines)}</blockquote>')
+        elif p.startswith('####'):
+            html_parts.append(f'<h4>{p.lstrip("# ").strip()}</h4>')
+        elif p.startswith('###'):
+            html_parts.append(f'<h3>{p.lstrip("# ").strip()}</h3>')
+        else:
+            p = re.sub(r'\*(.+?)\*', r'<em>\1</em>', p)
+            p = p.replace('\n', '<br>')
+            html_parts.append(f'<p>{p}</p>')
+
+    chapters.append({
+        "number": i,
+        "title": title,
+        "html": "\n".join(html_parts)
+    })
+
+out = website_dir / "chapters.json"
+out.write_text(json.dumps(chapters, indent=2))
+print(f"Wrote {len(chapters)} chapters ({out.stat().st_size / 1024:.1f} KB) to {out}")
--- a/website/chapters.json
+++ b/website/chapters.json
Author	SHA1	Message	Date
Alexander Whitestone	aaf257ad5f	feat: enhanced build system — self-contained HTML web reader, deploy version, reportlab PDF with QR codes - web-style.css: dark/light mode, responsive, EB Garamond typography - HTML build: self-contained (embedded CSS) + deploy version (external CSS) - Reading progress bar and back-to-top button injected via JS - Reader footer with author credit and 988 crisis line - All outputs consolidated in build/output/ - EPUB: 68KB, PDF: 143KB (reportlab), HTML: 3.7MB (self-contained)	2026-04-10 23:52:00 -04:00
Alexander Whitestone	2acc538fc4	wip: add web reader stylesheet with dark mode, progress bar, responsive design	2026-04-10 23:51:21 -04:00
Google AI Agent	3247dd29f0	[Testament] Add scripts/guardrails.sh (GOFAI improvements and guardrails) Some checks failed Smoke Test / smoke (push) Failing after 5s Details	2026-04-11 01:40:37 +00:00
Google AI Agent	a01b998f61	[Testament] Add build/semantic_linker.py (GOFAI improvements and guardrails) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:40:36 +00:00
Google AI Agent	59cd71985d	Create website/chapters.json Some checks failed Smoke Test / smoke (push) Failing after 5s Details	2026-04-11 01:36:33 +00:00
Google AI Agent	6c506caac6	Create website/build-chapters.py Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:32 +00:00
Google AI Agent	55d51f2ee4	Update testament-complete.md (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:31 +00:00
Google AI Agent	eae9398fa5	Update compile.py (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:29 +00:00
Google AI Agent	f8528e9ded	Update build/metadata.yaml (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:28 +00:00
Google AI Agent	374d82a886	Update build/frontmatter.md (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:27 +00:00
Google AI Agent	4763311588	Update build/build.py (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:25 +00:00
Google AI Agent	348ed7ee92	Update build/backmatter.md (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:24 +00:00
Google AI Agent	22f59c57cb	Create book-style.css Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:23 +00:00
Google AI Agent	4ac38f1b60	Create art-manifest.md Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:22 +00:00
Google AI Agent	d586fb211d	Update Makefile (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:20 +00:00
Google AI Agent	92867808b2	Update MULTIMEDIA-PLAN.md (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:18 +00:00
Google AI Agent	47a13325cc	Create EPIC-MATRIX.md Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:17 +00:00
Google AI Agent	14273702ba	Update .gitignore (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:16 +00:00
Google AI Agent	2e1f6ffb5b	Update .gitea/workflows/smoke.yml (manual merge) Some checks failed Smoke Test / smoke (push) Has been cancelled Details	2026-04-11 01:36:15 +00:00
Alexander Whitestone	08233364ff	burn: add smoke test workflow — parse check + secret scan All checks were successful Smoke Test / smoke-test (pull_request) Successful in 10s Details Build Validation / validate-manuscript (pull_request) Successful in 9s Details Smoke Test / smoke-test (push) Successful in 8s Details Closes #27 Adds a dead-simple CI smoke test that runs on every PR and push to main: Parse checks: - Chapter validation (structure, numbering, H1 headers) - Markdown build (combines all chapters) - Compiled manuscript size verification (>10k words) - Python syntax check on all .py files - YAML syntax check on workflow files Secret scan: - Scans for common API key/token patterns (sk-ant-, sk-or-, ghp_, AKIA, etc.) - Searches all text files, excludes .git and the smoke test itself - Hard fail if any secrets found Two files: - scripts/smoke.sh — the smoke test script - .gitea/workflows/smoke.yml — Gitea Actions workflow	2026-04-10 20:58:16 -04:00
Timmy Time	544bc1a985	Merge pull request 'feat: add CI workflow for manuscript build validation' (#25 ) from feat/ci-build-validation into main Merged PR #25: feat: add CI workflow for manuscript build validation	2026-04-11 00:44:01 +00:00
Timmy Time	ba9fd0ba08	Merge pull request 'burn: add chapter validation to build pipeline (closes #24 )' (#26 ) from burn/20260410-chapter-validation into main Merged PR #26: burn: add chapter validation to build pipeline	2026-04-11 00:43:38 +00:00
Timmy Time	8ba9f58e96	Merge pull request 'feat: add book compilation pipeline (rescued from #20 )' (#28 ) from rescue/book-compilation into main Merged PR #28: feat: add book compilation pipeline	2026-04-11 00:43:36 +00:00
Alexander Whitestone	f6d74e233b	feat: add book compilation pipeline (rescued from #20 ) Build system for The Testament: - build/build.py: compiles chapters to PDF, EPUB, MD - build/metadata.yaml: book metadata - build/frontmatter.md: title page, dedication - build/backmatter.md: acknowledgments, sovereignty note - Makefile: make pdf, make epub, make md - .gitignore: build artifacts	2026-04-10 20:32:38 -04:00
Alexander Whitestone	948d520b83	burn: add chapter validation to build pipeline (closes #24 ) Add validate_chapters() function that checks: - No empty chapter files (whitespace-only counts as empty) - Every chapter starts with an H1 header (# Chapter N — Title) - No gaps in chapter numbering (sequential from 1) - No duplicate chapter numbers - Header chapter number matches filename number - Warns on suspiciously short chapters (<50 words) Validation runs automatically before compilation. If errors are found, compilation is aborted with clear error messages showing exactly what to fix. CLI flags: python3 compile.py --validate # validate only python3 compile.py --no-validate # skip validation python3 compile.py # validate then compile	2026-04-10 19:57:27 -04:00
Google AI Agent	7a56b4b727	feat: add CI workflow for manuscript build validation Some checks failed Build Validation / validate-manuscript (pull_request) Failing after 5s Details	2026-04-10 23:55:17 +00:00
Timmy Time	bebd3943d4	[auto-merge] README update Auto-merged by PR review bot: README update	2026-04-10 11:48:32 +00:00
Alexander Whitestone	1d4e8a6478	burn: update README with full 18-chapter structure, characters, themes Closes #21 The README previously listed only Chapter 1 with 'Draft' status. Now includes: - All 18 chapters organized by part (I-V) - Status indicators with checkmark for Part I (complete) - Word count target (~70K) and current draft (~19K) - File inventory of repo contents - Character table with main cast - Core themes list from OUTLINE.md - Link to compilation pipeline PR #20	2026-04-10 06:42:59 -04:00
Alexander Whitestone	d0680715ac	Merge pull request #19 Merged PR #19	2026-04-10 03:43:49 +00:00