taketomarket 2.2.0 → 2.3.0

package/workflows/setup/init-validation.md

@@ -153,3 +153,25 @@ Relevant commands for warn follow-ups:
 - Warning flag format: Add `<!-- SPECIFICITY_WARNING: [field] accepted after 2 retries -->` comment in generated file
 - After accepting with warning, continue to next question (do not block the interview)
 - Track total warnings per section; if a section has 3+ warnings, note in STATE.md for follow-up
+
+---
+
+## PRODUCT-DNA validation
+
+After Section "PRODUCT-DNA":
+- **Beliefs check:** require at least 3 distinct beliefs, each containing a specific claim (not "we care about X"). FAIL if any belief is a generic value statement.
+- **Worldview check:** minimum 80 words. FAIL if shorter. FAIL if it contains marketing buzzwords ("revolutionary", "innovative", "disruptive", "best-in-class") - push for specific motivation instead.
+- **JTBD check:** primary JTBD must be a verb-led, customer-centric phrase. FAIL "be the best at X" or "leader in Y".
+
+## Brand colors validation
+
+After Section "Brand Colors":
+- **Primary color contrast:** primary must pass WCAG AA on either light or dark backgrounds. FAIL if neither.
+- **Hex validity:** all colors must be valid hex codes (3, 6, or 8 chars). FAIL invalid.
+
+## Logo validation
+
+After Section "Logo":
+- **File presence:** all 8 logo files must exist in .taketomarket/brand/. FAIL missing.
+- **SVG validity:** each .svg file must parse as valid XML. FAIL malformed.
+- **Vision-review history:** BRAND.md `## Logo` section must include a non-empty vision-review history. FAIL empty.

package/workflows/setup/init.md

@@ -1,8 +1,39 @@
+## Step 0: First-run inline education
+
+Read `.taketomarket/CONFIG.md`. Parse `first_run_seen` (object) and `inline_education` (boolean, default true).
+
+If `inline_education` is false: skip this step. Else if `first_run_seen.ttm-init` is not `true`, print the explainer below verbatim, then mark this skill as seen:
+
+```bash
+node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run mark ttm-init
+```
+
+Use this exact check (bash) to decide whether to print: `node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run check ttm-init --raw` -- the JSON `seen` field is `true` once the explainer has run before.
+
+### Explainer for `/ttm-init`
+
+`/ttm-init` is the bootstrap interview. It asks you ~30 structured questions
+about your product, audience, positioning, and brand, then generates the
+`.taketomarket/` reference files (POSITIONING.md, ICP.md, BRAND.md, PRODUCT-DNA.md,
+plus brand colors and a logo set) and writes CLAUDE.md / AGENTS.md so the runtime
+knows the rules. Think of it as `npm init` for marketing: one interactive pass
+produces the config that every other `/ttm-*` skill reads.
+
+Why it matters now: every downstream skill -- produce, verify, ship, measure --
+loads these reference files as their context. If you skip init, every later
+skill is operating without your positioning invariant and will drift. Run init
+once per project; re-run only on a real positioning shift via `/ttm-positioning-shift`.
+
+(Canonical source: `references/inline-education-blurbs.md`. Embedded verbatim because workflows do not @-resolve files at runtime.)
+
+---
+
 <purpose>
-Interview-driven onboarding that generates all .marketing/ reference files
+Interview-driven onboarding that generates all .taketomarket/ reference files
 from structured questioning. Use when setting up takeToMarket for a new project.
 Orchestrates a 6-section interview with specificity validation, then generates
-9 reference files plus CLAUDE.md and AGENTS.md instruction files.
+10 reference files (including PRODUCT-DNA) plus brand assets (colors.json + logo
+set) plus CLAUDE.md and AGENTS.md instruction files.
 </purpose>
 
 <required_reading>
@@ -40,6 +71,19 @@ For multiSelect questions, instruct the user: "Type the numbers of your choices
 
 ---
 
+## YOLO Mode Detection
+
+Read `.taketomarket/CONFIG.md` if it exists:
+```bash
+node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs config read --raw 2>/dev/null
+```
+
+If `yolo: true`: set `YOLO_MODE=true`. Throughout this workflow, when a question is tagged `priority: non-critical` in init-questions.md, skip it and use its documented `default:` value.
+
+Critical questions (positioning, manifesto beliefs/worldview, primary brand color, logo final approval) are NEVER skipped regardless of YOLO mode.
+
+---
+
 ## Step 1: Pre-flight
 
 ```
@@ -55,7 +99,7 @@ node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs init --raw
 
 Use AskUserQuestion:
 - header: "Existing Setup"
-- question: ".marketing/ already contains reference files. What would you like to do?"
+- question: ".taketomarket/ already contains reference files. What would you like to do?"
 - options:
   - label: "Start fresh"
     description: "Delete existing files and re-run the full interview"
@@ -64,14 +108,14 @@ Use AskUserQuestion:
   - label: "Cancel"
     description: "Exit without changes"
 
-**If "Start fresh":** Delete .marketing/ contents (except CAMPAIGNS/ and PLAYBOOKS/) and continue to Step 2.
-**If "Update specific files":** Present file picker using AskUserQuestion with multiSelect: true listing all 9 reference files. Run only the interview sections that map to selected files (see Question-to-Template Mapping in init-questions.md). Skip unselected sections and jump to Step 9 for generation.
+**If "Start fresh":** Delete .taketomarket/ contents (except CAMPAIGNS/ and PLAYBOOKS/) and continue to Step 2.
+**If "Update specific files":** Present file picker using AskUserQuestion with multiSelect: true listing all 10 reference files (POSITIONING, BRAND, ICP, CHANNELS, COMPETITORS, METRICS, CALENDAR, STATE, LEARNINGS, PRODUCT-DNA). Run only the interview sections that map to selected files (see Question-to-Template Mapping in init-questions.md). If PRODUCT-DNA is selected, re-run the sub-workflow at `${CLAUDE_PLUGIN_ROOT}/workflows/setup/init-product-dna.md`. Skip unselected sections and jump to Step 9 for generation.
 **If "Cancel":** Exit with message "No changes made."
 
 **If result is "not initialized":**
 ```bash
-mkdir -p .marketing/CAMPAIGNS
-mkdir -p .marketing/PLAYBOOKS
+mkdir -p .taketomarket/CAMPAIGNS
+mkdir -p .taketomarket/PLAYBOOKS
 ```
 Continue to Step 2.
 
@@ -176,6 +220,42 @@ On WARN: "Your customer language could be more detailed, but I'll accept it for
 
 ---
 
+## Step 4b: PRODUCT-DNA Generation
+
+```
+takeToMarket > PRODUCT-DNA (WHAT + WHY)
+```
+
+Read and follow `${CLAUDE_PLUGIN_ROOT}/workflows/setup/init-product-dna.md`.
+
+This sub-workflow scans the codebase (WHAT) and interviews the user for manifesto + worldview (WHY), then writes `.taketomarket/PRODUCT-DNA.md`. It depends on POSITIONING + BRAND + ICP for context coherence and must run before brand colors + logo (which use PRODUCT-DNA's worldview for rationale generation).
+
+---
+
+## Step 4c: Brand Colors
+
+```
+takeToMarket > BRAND COLORS
+```
+
+Read and follow `${CLAUDE_PLUGIN_ROOT}/workflows/setup/init-brand-colors.md`.
+
+This sub-workflow proposes 3 WCAG-checked palettes, lets the user pick, then appends colors to BRAND.md and saves `.taketomarket/brand/colors.json` for landing/pSEO consumption.
+
+---
+
+## Step 4d: Logo
+
+```
+takeToMarket > LOGO
+```
+
+Read and follow `${CLAUDE_PLUGIN_ROOT}/workflows/setup/init-logo.md`.
+
+This sub-workflow generates SVG candidates with vision self-review (up to 3 rounds), produces a full asset set under `.taketomarket/brand/`, and appends a Logo section to BRAND.md.
+
+---
+
 ## Step 5: Interview -- Section 4: Channels
 
 ```
@@ -261,7 +341,7 @@ Use AskUserQuestion:
 - question: "Here's what I'll use to generate your reference files. Ready to proceed?"
 - options:
   - label: "Generate files"
-    description: "Create all .marketing/ reference files"
+    description: "Create all .taketomarket/ reference files"
   - label: "Revise a section"
     description: "Go back and update specific answers"
 
@@ -286,44 +366,44 @@ For each reference file, follow this pattern:
    - Summary section must be under 200 words
    - Preserve all table structures from the template
    - For dates, use: `node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs timestamp date --raw`
-4. Write to `.marketing/{FILENAME}.md`
+4. Write to `.taketomarket/{FILENAME}.md`
 
 **Generate files in this order:**
 
 ### 1. POSITIONING.md (from Section 1 data)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/positioning.md`
 Fill: Category, Target audience, Primary differentiator, Proof points, Must-not-say, Competitive frame, Positioning History date.
-Write to: `.marketing/POSITIONING.md`
+Write to: `.taketomarket/POSITIONING.md`
 
 ### 2. BRAND.md (from Section 2 data)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/brand.md`
 Fill: Voice archetype, Voice attributes, Tone per context, Banned words with reasoning, Proof points, Good and bad examples.
-Write to: `.marketing/BRAND.md`
+Write to: `.taketomarket/BRAND.md`
 
 ### 3. ICP.md (from Section 3 data)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/icp.md`
 Fill: Primary segment, Demographics, Psychographics, JTBD, Pain points with severity and frequency, Buying triggers, Anti-ICP, Customer language library.
-Write to: `.marketing/ICP.md`
+Write to: `.taketomarket/ICP.md`
 
 ### 4. CHANNELS.md (from Section 4 data)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/channels.md`
 Fill: Active channels with baselines, Primary channel, Dormant channels with reasons, Banned channels with reasons, Budget allocation, Channel-specific rules.
-Write to: `.marketing/CHANNELS.md`
+Write to: `.taketomarket/CHANNELS.md`
 
 ### 5. COMPETITORS.md (from Section 5 data)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/competitors.md`
 Fill: Direct competitors (positioning, strength, weakness), Positioning map axes and positions, Share of voice baseline.
-Write to: `.marketing/COMPETITORS.md`
+Write to: `.taketomarket/COMPETITORS.md`
 
 ### 6. METRICS.md (from Section 6 data -- metrics portion)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/metrics.md`
 Fill: Primary outcome metric (metric/target/window/source), Secondary metrics, Leading indicators, Baselines, Attribution model.
-Write to: `.marketing/METRICS.md`
+Write to: `.taketomarket/METRICS.md`
 
 ### 7. CALENDAR.md (from Section 6 data -- calendar portion)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/calendar.md`
 Fill: Quarterly themes, Launch calendar, Always-on cadence, Blackout dates.
-Write to: `.marketing/CALENDAR.md`
+Write to: `.taketomarket/CALENDAR.md`
 
 ### 8. STATE.md (template copy with timestamp)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/state.md`
@@ -332,7 +412,7 @@ Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/state.md`
 node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs timestamp date --raw
 ```
 Do NOT fill interview data into STATE.md -- it is initialized with default values.
-Write to: `.marketing/STATE.md`
+Write to: `.taketomarket/STATE.md`
 
 ### 9. LEARNINGS.md (template copy with zeroed counters)
 Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/learnings.md`
@@ -341,7 +421,7 @@ Read template: `${CLAUDE_PLUGIN_ROOT}/templates/reference-files/learnings.md`
 - Last lesson date: none
 - Top pattern: none
 Do NOT fill interview data -- learnings are populated during campaign Learn phases.
-Write to: `.marketing/LEARNINGS.md`
+Write to: `.taketomarket/LEARNINGS.md`
 
 ---
 
@@ -384,7 +464,7 @@ node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs state update status initialized
 ```
 
 **Warning tracking:** If any interview section accumulated 3 or more SPECIFICITY_WARNING flags
-during the interview, append to `.marketing/STATE.md` under a `## Follow-up Needed` heading:
+during the interview, append to `.taketomarket/STATE.md` under a `## Follow-up Needed` heading:
 - Section [N] ([name]): [count] specificity warnings -- recommend re-running with [relevant /ttm-* command]
 
 Use these commands per section:
@@ -403,15 +483,27 @@ Use these commands per section:
 takeToMarket > INITIALIZED
 
 Reference files created:
-  .marketing/POSITIONING.md
-  .marketing/BRAND.md
-  .marketing/ICP.md
-  .marketing/CHANNELS.md
-  .marketing/STATE.md
-  .marketing/METRICS.md
-  .marketing/COMPETITORS.md
-  .marketing/CALENDAR.md
-  .marketing/LEARNINGS.md
+  .taketomarket/POSITIONING.md
+  .taketomarket/BRAND.md
+  .taketomarket/ICP.md
+  .taketomarket/CHANNELS.md
+  .taketomarket/STATE.md
+  .taketomarket/METRICS.md
+  .taketomarket/COMPETITORS.md
+  .taketomarket/CALENDAR.md
+  .taketomarket/LEARNINGS.md
+  .taketomarket/PRODUCT-DNA.md
+
+Brand assets created:
+  .taketomarket/brand/colors.json
+  .taketomarket/brand/logo.svg
+  .taketomarket/brand/logo-mark.svg
+  .taketomarket/brand/logo-wordmark.svg
+  .taketomarket/brand/logo-mono-light.svg
+  .taketomarket/brand/logo-mono-dark.svg
+  .taketomarket/brand/logo-favicon.svg
+  .taketomarket/brand/logo@1x.png
+  .taketomarket/brand/logo@2x.png
 
 Instruction files created:
   CLAUDE.md
@@ -427,23 +519,36 @@ Next step: Run /ttm-new-campaign to create your first campaign.
 <success_criteria>
 - [ ] Pre-flight check completed (existing init detected or directories created)
 - [ ] All 6 interview sections completed with specificity validation
+- [ ] PRODUCT-DNA sub-workflow run; `.taketomarket/PRODUCT-DNA.md` written
+- [ ] Brand-colors sub-workflow run; BRAND.md `## Colors` section filled + `.taketomarket/brand/colors.json` written
+- [ ] Logo sub-workflow run; `.taketomarket/brand/` asset set produced + BRAND.md `## Logo` section filled
 - [ ] Confirmation gate passed
-- [ ] 9 reference files generated in .marketing/ with _SUMMARY/END_SUMMARY preserved
+- [ ] 10 reference files generated in .taketomarket/ with _SUMMARY/END_SUMMARY preserved
 - [ ] CLAUDE.md and AGENTS.md copied to project root
-- [ ] Health check passes (all 11 checks: marketing_dir + campaigns_dir + 9 files)
+- [ ] Health check passes (all 12 checks: taketomarket_dir + campaigns_dir + 10 files)
 - [ ] STATE.md status set to "initialized"
 </success_criteria>
 
 <output>
-- `.marketing/POSITIONING.md`
-- `.marketing/BRAND.md`
-- `.marketing/ICP.md`
-- `.marketing/CHANNELS.md`
-- `.marketing/STATE.md`
-- `.marketing/METRICS.md`
-- `.marketing/COMPETITORS.md`
-- `.marketing/CALENDAR.md`
-- `.marketing/LEARNINGS.md`
+- `.taketomarket/POSITIONING.md`
+- `.taketomarket/BRAND.md`
+- `.taketomarket/ICP.md`
+- `.taketomarket/CHANNELS.md`
+- `.taketomarket/STATE.md`
+- `.taketomarket/METRICS.md`
+- `.taketomarket/COMPETITORS.md`
+- `.taketomarket/CALENDAR.md`
+- `.taketomarket/LEARNINGS.md`
+- `.taketomarket/PRODUCT-DNA.md`
+- `.taketomarket/brand/colors.json`
+- `.taketomarket/brand/logo.svg` (plus mark/wordmark/mono/favicon variants and PNG raster)
 - `CLAUDE.md`
 - `AGENTS.md`
 </output>
+
+## What if this doesn't fit?
+
+Looks like /ttm-init can't do that yet.
+
+- Want a new skill? /ttm-request-skill
+- Existing skill needs work? /ttm-improve-skill

package/workflows/setup/new-campaign.md

@@ -1,3 +1,32 @@
+## Step 0: First-run inline education
+
+Read `.taketomarket/CONFIG.md`. Parse `first_run_seen` (object) and `inline_education` (boolean, default true).
+
+If `inline_education` is false: skip this step. Else if `first_run_seen.ttm-new-campaign` is not `true`, print the explainer below verbatim, then mark this skill as seen:
+
+```bash
+node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run mark ttm-new-campaign
+```
+
+Use this exact check (bash) to decide whether to print: `node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run check ttm-new-campaign --raw` -- the JSON `seen` field is `true` once the explainer has run before.
+
+### Explainer for `/ttm-new-campaign`
+
+`/ttm-new-campaign` creates an isolated workspace under `.taketomarket/campaigns/<slug>/`
+with its own STATE.md, brief slot, and asset directory. It's the marketing
+equivalent of `git checkout -b feature/X`: campaigns are independent units that
+can be at different lifecycle phases simultaneously, and the state machine
+tracks each one separately.
+
+Why it matters: takeToMarket is multi-campaign by design. Without a campaign
+slug, briefs, assets, and verification results have no home and bleed into
+each other. Run this before `/ttm-brief` for any new initiative -- it locks in
+the campaign id that subsequent skills route work into.
+
+(Canonical source: `references/inline-education-blurbs.md`. Embedded verbatim because workflows do not @-resolve files at runtime.)
+
+---
+
 <purpose>
 Campaign scaffolding workflow that creates a CAMPAIGNS/<slug>/ directory with
 initialized STATE.md, empty RESEARCH.md, empty BRIEF.md, and ASSETS/ directory.
@@ -52,7 +81,7 @@ Display: `Campaign slug: ${CAMPAIGN_SLUG}`
 ## Step 3: Check for Existing Campaign
 
 ```bash
-ls .marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/ 2>/dev/null && echo "exists" || echo "new"
+ls .taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/ 2>/dev/null && echo "exists" || echo "new"
 ```
 
 **If "exists":**
@@ -81,10 +110,10 @@ This creates:
 
 Verify all 4 items exist:
 ```bash
-ls .marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/STATE.md && \
-ls .marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/RESEARCH.md && \
-ls .marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/BRIEF.md && \
-ls -d .marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/ASSETS/ && \
+ls .taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/STATE.md && \
+ls .taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/RESEARCH.md && \
+ls .taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/BRIEF.md && \
+ls -d .taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/ASSETS/ && \
 echo "scaffold complete"
 ```
 
@@ -106,13 +135,13 @@ node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" state update current_campaign "${
 ```
 takeToMarket > CAMPAIGN CREATED: ${CAMPAIGN_SLUG}
 
-.marketing/CAMPAIGNS/${CAMPAIGN_SLUG}/
+.taketomarket/CAMPAIGNS/${CAMPAIGN_SLUG}/
   STATE.md       (phase: created)
-  RESEARCH.md    (pending -- run /ttm-research)
+  RESEARCH.md    (pending -- run /ttm-discover)
   BRIEF.md       (pending -- run /ttm-brief)
   ASSETS/        (empty -- populated by /ttm-produce)
 
-Next: Run /ttm-research ${CAMPAIGN_SLUG}
+Next: Run /ttm-discover ${CAMPAIGN_SLUG}
 ```
 
 </process>
@@ -127,8 +156,15 @@ Next: Run /ttm-research ${CAMPAIGN_SLUG}
 </success_criteria>
 
 <output>
-- `.marketing/CAMPAIGNS/<slug>/STATE.md`
-- `.marketing/CAMPAIGNS/<slug>/RESEARCH.md`
-- `.marketing/CAMPAIGNS/<slug>/BRIEF.md`
-- `.marketing/CAMPAIGNS/<slug>/ASSETS/` (empty directory)
+- `.taketomarket/CAMPAIGNS/<slug>/STATE.md`
+- `.taketomarket/CAMPAIGNS/<slug>/RESEARCH.md`
+- `.taketomarket/CAMPAIGNS/<slug>/BRIEF.md`
+- `.taketomarket/CAMPAIGNS/<slug>/ASSETS/` (empty directory)
 </output>
+
+## What if this doesn't fit?
+
+Looks like /ttm-new-campaign can't do that yet.
+
+- Want a new skill? /ttm-request-skill
+- Existing skill needs work? /ttm-improve-skill

package/workflows/site/deploy.md

@@ -0,0 +1,98 @@
+# Deploy Workflow
+
+## Step 0: First-run inline education
+
+Read `.taketomarket/CONFIG.md`. Parse `first_run_seen` (object) and `inline_education` (boolean, default true).
+
+If `inline_education` is false: skip this step. Else if `first_run_seen.ttm-deploy` is not `true`, print the explainer below verbatim, then mark this skill as seen:
+
+```bash
+node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run mark ttm-deploy
+```
+
+Use this exact check (bash) to decide whether to print: `node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run check ttm-deploy --raw` -- the JSON `seen` field is `true` once the explainer has run before.
+
+### Explainer for `/ttm-deploy`
+
+`/ttm-deploy` ships your generated site (landing + pSEO + any static
+assets) to the deploy target detected from your project (Vercel,
+Netlify, Cloudflare Pages, etc.) or guides manual deployment if no
+target is detected. It does not bypass the gate wall -- only verified,
+reviewed assets are pushed.
+
+Why it matters: a verified asset that isn't deployed is functionally
+equivalent to no asset. This skill is the bridge between the
+spec-driven pipeline and the live URL, and it preserves the audit trail
+so you can correlate a deployed page back to its brief and gate results.
+
+(Canonical source: `references/inline-education-blurbs.md`. Embedded verbatim because workflows do not @-resolve files at runtime.)
+
+---
+
+## Step 1: Read landing path
+
+```bash
+node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs config read --raw
+```
+
+Extract `landing_path`. If missing: print "Run /ttm-landing first" and exit.
+
+## Step 2: Detect deploy path
+
+```bash
+cd <landing_path> && node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs deploy detect --raw
+```
+
+Parse JSON. Branch on `preferred`:
+
+### preferred = "git-push"
+
+Inside landing path:
+1. `git add . && git commit -m "deploy: takeToMarket landing update"` (if changes).
+2. `git push origin <current-branch>`.
+3. Print Vercel dashboard URL and "Deploy in progress — check dashboard."
+
+### preferred = "cli"
+
+Inside landing path:
+1. `vercel deploy [--prod]` (depending on flag).
+2. CLI outputs deploy URL. Print it.
+
+### preferred = "api-token"
+
+Print: "API-token deploy not yet implemented for v2.3.0. Either install vercel CLI or connect this repo to Vercel and use git-push."
+
+### preferred = null (nothing available)
+
+Walk user through setup:
+1. AskUserQuestion: "No Vercel deploy path detected. Which would you like to set up?"
+   - "Connect repo to Vercel dashboard (recommended)" — print instructions.
+   - "Install Vercel CLI" — `npm i -g vercel && vercel login`.
+   - "Skip — I'll deploy manually."
+
+## Step 3: Verify deploy URL responds
+
+If a deploy URL was produced, fetch the homepage:
+```bash
+curl -sI <url> | head -1
+```
+Expected: `HTTP/2 200`.
+
+If non-200: print warning + URL.
+
+## Step 4: Update CONFIG.md
+
+Append:
+```
+last_deploy_url: <url>
+last_deploy_at: <timestamp>
+```
+
+## Step 5: Print next steps
+
+## What if this doesn't fit?
+
+Looks like /ttm-deploy can't do that yet.
+
+- Want a new skill? /ttm-request-skill
+- Existing skill needs work? /ttm-improve-skill

package/workflows/site/landing.md

@@ -0,0 +1,156 @@
+# Landing Page Workflow
+
+## Step 0: First-run inline education
+
+Read `.taketomarket/CONFIG.md`. Parse `first_run_seen` (object) and `inline_education` (boolean, default true).
+
+If `inline_education` is false: skip this step. Else if `first_run_seen.ttm-landing` is not `true`, print the explainer below verbatim, then mark this skill as seen:
+
+```bash
+node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run mark ttm-landing
+```
+
+Use this exact check (bash) to decide whether to print: `node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run check ttm-landing --raw` -- the JSON `seen` field is `true` once the explainer has run before.
+
+### Explainer for `/ttm-landing`
+
+`/ttm-landing` produces a landing page from your positioning, brand,
+ICP, and a brief. It generates the HTML, applies the brand tokens,
+writes the page into your site directory, and queues the four landing
+quality gates (positioning, copy clarity, conversion fundamentals,
+visual review via Playwright).
+
+Why it matters: landing pages are the highest-leverage marketing
+asset for developerneurs because they're the conversion surface for
+every other campaign. Treating them as one-off prose vs. spec-driven
+generation is the difference between A/B testing your way upward
+and rewriting from scratch every two months.
+
+(Canonical source: `references/inline-education-blurbs.md`. Embedded verbatim because workflows do not @-resolve files at runtime.)
+
+---
+
+**Required reading:**
+- `.taketomarket/POSITIONING.md`
+- `.taketomarket/BRAND.md`
+- `.taketomarket/PRODUCT-DNA.md`
+- `.taketomarket/ICP.md`
+- `${CLAUDE_PLUGIN_ROOT}/references/landing-page-anatomy.md`
+- `${CLAUDE_PLUGIN_ROOT}/playbooks/landing-pages.md` (delivered in P6)
+
+---
+
+## Step 0: Prerequisite check
+
+Verify `/ttm-init` (P3) has run by reading `.taketomarket/brand/colors.json`. If the file is missing:
+
+```
+Run /ttm-init first to generate POSITIONING, BRAND, PRODUCT-DNA, ICP and brand colors.
+```
+
+Exit. Do not proceed — without `colors.json`, the `{{COLOR_*}}` placeholders in `app/tokens.css` will not substitute and the scaffolded site will render unstyled.
+
+## Step 1: Choose site location
+
+Run:
+```bash
+node ${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs site-location --raw
+```
+
+Parse the JSON. Use AskUserQuestion (priority: critical):
+- question: "Where should the landing site live?"
+- options:
+  - "Default: [result.default]"
+  - "Custom path"
+  - "Cancel"
+
+If "Custom path": freeform input.
+
+Save chosen path to `.taketomarket/CONFIG.md` `landing_path: <path>`.
+
+## Step 2: Copy scaffold
+
+Copy `${CLAUDE_PLUGIN_ROOT}/templates/site-scaffold/` to chosen landing path. Use `cp -R` or fs equivalent.
+
+Skip if landing path already exists with a `package.json` (preserve user work).
+
+## Step 3: Template substitution
+
+Read `.taketomarket/brand/colors.json`. Replace `{{COLOR_*}}` placeholders in `app/tokens.css`.
+
+Read POSITIONING.md, BRAND.md. Replace `{{SITE_TITLE}}`, `{{SITE_DESCRIPTION}}`, `{{SITE_NAME}}` in `app/layout.tsx`, `package.json`, `public/llms.txt`.
+
+## Step 4: Generate copy for each section
+
+For each of 13 sections in `app/page.tsx`, generate content following references/landing-page-anatomy.md guidance. Load positioning differentiator, brand voice, PRODUCT-DNA worldview for context.
+
+Fill in component placeholders (`{{HERO_HEADLINE}}` etc.) with generated copy.
+
+## Step 4b: Inject Schema.org JSON-LD into each page
+
+Quality Gate 2 (`workflows/site/quality-gates.md`) requires a `<script type="application/ld+json">` block on every page. The scaffold ships without one — add it here.
+
+For each page generated under `app/`, inject a JSON-LD `<script>` into the page's JSX (top of the return tree is fine; Next.js will render it into the document head/body). Schema type to use per page (from `references/landing-page-anatomy.md`):
+
+- `app/page.tsx` (home): `Organization` + `WebSite` (combined in a single `@graph` array if both are needed).
+- `app/product/page.tsx`: `Product`.
+- `app/pricing/page.tsx`: `Product` with `offers` (or `Service` if SaaS).
+- `app/about/page.tsx`: `Organization` (extended with `founder`, `foundingDate`).
+- Every page with an FAQ section: add a `FAQPage` entry next to the page-type entry.
+
+Use values from POSITIONING.md (name, description, differentiator → about/description), BRAND.md (logo URL, social profiles → `sameAs`), and the generated copy itself (FAQ questions/answers → `FAQPage.mainEntity`).
+
+JSON-LD must be valid JSON. Render via `<script type="application/ld+json" dangerouslySetInnerHTML={{ __html: JSON.stringify(schema) }} />`.
+
+## Step 5: Generate /product, /pricing, /about pages
+
+Use AskUserQuestion (priority: non-critical) for which pages to scaffold:
+- "Which top-level pages do you want now?"
+- multiSelect: ["Home (always)", "Product", "Pricing", "About / Manifesto"]
+
+For each selected page: scaffold + generate copy following references/landing-page-anatomy.md.
+
+## Step 6: Mandatory humanize
+
+For every generated copy file, invoke `/ttm-humanize` via Skill tool. Replace original with humanized version.
+
+## Step 7: Initial commit in the landing project
+
+First detect whether the landing path is already inside a git work tree:
+
+```bash
+cd <landing_path> && git rev-parse --is-inside-work-tree 2>/dev/null
+```
+
+- If the command prints `true`: the landing path is inside an existing repo (parent project). Do NOT `git init` — that would nest repositories. Just `git add . && git commit -m "feat: takeToMarket landing scaffold + initial copy"` at the parent repo level.
+- If the command exits non-zero (not a repo): initialize a fresh repo:
+
+```bash
+cd <landing_path> && git init && git add . && git commit -m "feat: takeToMarket landing scaffold + initial copy"
+```
+
+## Step 8: Quality gates
+
+Run quality gates from `${CLAUDE_PLUGIN_ROOT}/workflows/site/quality-gates.md`:
+- Positioning integrity (parse rendered HTML, verify differentiator + must-not-say compliance).
+- Banned words check.
+- (Playwright visual + perf — soft until P5 ships Playwright setup.)
+
+If gates fail: route to /ttm-fix.
+
+## Step 9: Print next steps
+
+```
+✓ Landing site scaffolded at [path].
+Next:
+- Run npm install && npm run dev in [path] to preview.
+- Run /ttm-deploy when ready to ship.
+- Run /ttm-pseo to add blog/use-case/comparison/alternative routes.
+```
+
+## What if this doesn't fit?
+
+Looks like /ttm-landing can't do that yet.
+
+- Want a new skill? /ttm-request-skill
+- Existing skill needs work? /ttm-improve-skill