mattgierhart · mattgierhart · May 31, 2026 · May 22, 2026 · May 22, 2026 · May 22, 2026
diff --git a/.claude/agents/metro/MEMORY.md b/.claude/agents/metro/MEMORY.md
@@ -12,5 +12,14 @@
 ## Decisions (choices made with rationale)
 <!-- Format: Decision. **Why:** rationale. **Alternatives rejected:** what and why -->
 
+**v0.9 GTM is now a chained framework workflow, not a monolithic skill.** Run order: `prd-v09-positioning-dunford` (Dunford 5-step) → `prd-v09-offer-construction-hormozi` ($100M Offers value equation) → `prd-v09-launch-channels-orb` (Owned/Rented/Borrowed). `prd-v09-gtm-strategy` is now the orchestrator that runs these three and reconciles outputs (best-fit alignment, promise consistency, guarantee viability, category-channel fit, KPI rollup). **Why:** monolithic skill was too broad and produced incoherent output across positioning/offer/channels. **Alternatives rejected:** keep monolithic with deeper sub-sections (still too long; no framework grounding); fully delete old skill (breaks backwards compat with existing GTM- entries).
+
+**v1.0 is anchored on Moore's Crossing the Chasm with new ADO- ID prefix.** Stage assessment (ADO-STAGE-), beachhead segment (ADO-BEACHHEAD-), whole-product gaps (ADO-WHOLE-), reference accounts (ADO-REF-) all live in `SoT/SoT.ADOPTION.md`. **Why:** v1.0 was previously a stub; Moore's lifecycle is the canonical framework for post-launch adoption strategy and pragmatist-buyer behavior. **Alternatives rejected:** keep v1.0 as a PRD.md inline section without first-class SoT (no queryable adoption data; weak traceability for chasm-crossing risks).
+
+**Tactical playbooks attach to channels from the ORB matrix, not to GTM-strategy directly.** AEO audit, alternatives pages, cold outreach tiered, HN/Reddit launch all run after Launch Channels (ORB) has produced its channel-mix matrix. Each playbook explicitly disqualifies itself when positioning best-fit doesn't match the channel (e.g., HN/Reddit skips for enterprise procurement products). **Why:** running tactical playbooks before strategic channel selection produces channel work that doesn't reach the right segment.
+
+**Mode tiering (quick/standard/deep) applies to all v0.8/v0.9/v1.0 skills.** Default is standard; quick is appropriate (not degraded). See `.claude/rules/08-skill-execution-modes.md` for full doctrine. **Why:** same operation often needs 15-min founder pass OR 90-min deep audit; previously the only mode was implicitly "standard" which forced over-work on simple cases.
+
+
 ## Handoff Notes (friction points between agents)
 <!-- Format: From→To. **Issue:** what broke. **Fix:** what works -->
diff --git a/.claude/domain-profile.yaml b/.claude/domain-profile.yaml
@@ -37,11 +37,12 @@ id_prefixes:
   ARC:  { file: "SoT/SoT.TECHNICAL_DECISIONS.md",   description: "Architecture decisions" }
   INT:  { file: "SoT/SoT.INTEGRATIONS.md",          description: "Integrations" }
   LL:   { file: "SoT/SoT.LESSONS_LEARNED.md",       description: "Lessons learned (cross-session feedback)" }
+  ADO:  { file: "SoT/SoT.ADOPTION.md",               description: "Adoption-stage data (Moore Chasm — stage, beachhead, whole-product, references)" }
 
   # PRD/README-owned IDs
   FEA:  { file: "PRD.md",    description: "Features (PRD-owned, v0.3)" }
   RISK: { file: "PRD.md",    description: "Risks (PRD-owned, v0.5)" }
-  GTM:  { file: "PRD.md",    description: "Go-to-market (PRD-owned, v0.9)" }
+  GTM:  { file: "PRD.md",    description: "Go-to-market (PRD-owned, v0.9 — Type sub-codes: Positioning, Offer, Guarantee, Channel, Sequence, Index, AEO, OUT, CHG, MOPS, CASE, TST)" }
   KPI:  { file: "README.md", description: "KPI metrics (README-owned)" }
 
   # Execution IDs
@@ -81,9 +82,24 @@ skills:
     - prd-v08-monitoring-setup
     - prd-v08-release-planning
     - prd-v08-runbook-creation
+    - prd-v08-changelog-as-marketing
+    - prd-v08-drift-baseline-compare
+    - prd-v08-marketing-ops-handoff
     - prd-v09-feedback-loop-setup
     - prd-v09-gtm-strategy
     - prd-v09-launch-metrics
+    - prd-v09-positioning-dunford
+    - prd-v09-offer-construction-hormozi
+    - prd-v09-launch-channels-orb
+    - prd-v09-aeo-audit
+    - prd-v09-alternatives-pages
+    - prd-v09-cold-outreach-tiered
+    - prd-v09-hn-reddit-launch
+    - prd-v10-chasm-adoption-moore
+    - prd-v10-continuous-discovery-torres
+    - prd-v10-mom-test-interview
+    - prd-v10-case-study-builder
+    - prd-v10-testimonial-collector
 
 # --- Agent Registry ---
 # Product-lifecycle agents. Non-product repos may drop these entirely.

diff --git a/.claude/rules/08-skill-execution-modes.md b/.claude/rules/08-skill-execution-modes.md
@@ -0,0 +1,73 @@
+# Skill Execution Modes (quick / standard / deep)
+
+PRD-lifecycle skills run in one of three depth modes. The mode shapes time budget, output depth, and which optional steps run — it does not change whether the skill is correct or which Source-of-Truth IDs it owns.
+
+| Mode | Time | When to pick it | Output shape |
+|------|------|-----------------|--------------|
+| **quick** | < 15 min | Solo founder gut-check, "before I commit", rough first pass | 3–5 SoT IDs, no validation rounds, confidence may be 1/5 |
+| **standard** (default) | 30–45 min | Normal session, first real attempt at the stage | Full skill execution, every `Consumes` input checked, confidence ≥ 2/5 |
+| **deep** | 60–90 min | Critical decision, team-facing, pre-investor, audit | Full + research loops + assumptions log + cross-skill validation, confidence ≥ 3/5 where evidence exists |
+
+## Picking a mode
+
+Default is **standard**. Use the user's framing to infer otherwise.
+
+| Signal in user request | Mode |
+|------------------------|------|
+| "rough idea", "quick pass", "gut-check", "before I commit" | quick |
+| "define", "set up", "plan", "let's do" | standard |
+| "audit", "deep dive", "thorough", "investor-ready", "team review", "stress-test" | deep |
+
+Quick is **appropriate, not degraded**. Don't refuse it because "the proper way is more thorough" — the user picked the trade. Honor the time budget by cutting optional steps (`[standard+]`, `[deep only]`), not by speed-running every step.
+
+Only ask the user when framing is genuinely ambiguous.
+
+## Tagging skill phases (for skill authors)
+
+In `SKILL.md` execution steps, annotate optional phases:
+
+- *(no tag)* — runs in all modes
+- `[standard+]` — runs in standard and deep only
+- `[deep only]` — runs in deep only
+
+Each skill's frontmatter declares supported modes:
+
+```yaml
+execution_modes:
+  default: standard
+  supports: [quick, standard, deep]
+```
+
+If a skill genuinely only supports one mode (e.g., a `TEST-` generator that must be exhaustive), declare `supports: [standard]` and explain why in the SKILL.md preamble.
+
+## How depth shows up in outputs
+
+| Output dimension | quick | standard | deep |
+|------------------|-------|----------|------|
+| SoT IDs created | 3–5 | full | full + assumption log |
+| Confidence floor ([P4](../skills/PRINCIPLES.md)) | 1/5 OK | 2/5 minimum | 3/5 minimum where evidence exists |
+| Consumes / Produces sections | always emitted | always emitted | always emitted |
+| Consumes detail | only obviously-linked IDs | full chain | full chain + downstream impact analysis |
+| Anti-pattern check | skipped | done | done with concrete examples called out |
+| Cross-skill validation | none | one pass | iterative until clean |
+
+`Consumes` and `Produces` sections are emitted in **every** mode — the mode affects depth of *content*, not whether the section exists.
+
+## Anti-patterns
+
+| Pattern | Fix |
+|---------|-----|
+| Always running deep regardless of question | Pick the mode that fits the decision |
+| Calling output "quick" but doing standard work | Honor the budget; cut optional phases |
+| Skipping confidence floors in deep mode | Deep raises the evidence bar, doesn't lower it |
+| Adding `[deep only]` to mandatory phases | If a phase is required for correctness, it runs in all modes |
+| Quick mode without acknowledging the trade | Say which optional phases you skipped and why |
+
+## Relationship to P3 and P4
+
+Modes are a **time/scope dial**, not a research bypass:
+
+- **quick mode** acknowledges confidence will be 1/5–2/5 and tags outputs accordingly. It does not pretend evidence is stronger than it is.
+- **deep mode** raises the confidence floor — if you can't reach 3/5 with available evidence, the skill output should explicitly say *"blocked on research"* and surface the missing inputs.
+
+See [`PRINCIPLES.md`](../skills/PRINCIPLES.md) P3 (Research Drives Scope) and P4 (SoT is Living Evidence) for the underlying discipline.
diff --git a/.claude/skills/README.md b/.claude/skills/README.md
@@ -52,17 +52,32 @@ skills/
 ├── prd-v08-release-planning/              # v0.8 Release
 ├── prd-v08-runbook-creation/              # v0.8 Release
 ├── prd-v08-monitoring-setup/              # v0.8 Release
-├── prd-v09-gtm-strategy/                  # v0.9 Launch
+├── prd-v08-changelog-as-marketing/        # v0.8 Release (bridge to GTM)
+├── prd-v08-drift-baseline-compare/        # v0.8 Release (drift monitoring)
+├── prd-v08-marketing-ops-handoff/         # v0.8 Release (lead lifecycle)
+├── prd-v09-gtm-strategy/                  # v0.9 Launch (orchestrator)
+├── prd-v09-positioning-dunford/           # v0.9 Launch (Dunford 5-step)
+├── prd-v09-offer-construction-hormozi/    # v0.9 Launch (Hormozi $100M Offers)
+├── prd-v09-launch-channels-orb/           # v0.9 Launch (ORB framework)
+├── prd-v09-aeo-audit/                     # v0.9 Launch (AI search visibility)
+├── prd-v09-alternatives-pages/            # v0.9 Launch (X-vs-Y SEO pages)
+├── prd-v09-cold-outreach-tiered/          # v0.9 Launch (Tier 1/2/3 outreach)
+├── prd-v09-hn-reddit-launch/              # v0.9 Launch (HN + Reddit playbook)
 ├── prd-v09-launch-metrics/                # v0.9 Launch
 ├── prd-v09-feedback-loop-setup/           # v0.9 Launch
+├── prd-v10-chasm-adoption-moore/          # v1.0 Adoption (Moore — the spine)
+├── prd-v10-continuous-discovery-torres/   # v1.0 Adoption (Torres)
+├── prd-v10-mom-test-interview/            # v1.0 Adoption (Fitzpatrick)
+├── prd-v10-case-study-builder/            # v1.0 Adoption (social proof)
+├── prd-v10-testimonial-collector/         # v1.0 Adoption (social proof)
 ├── ghm-gate-check/                        # Methodology
 ├── ghm-id-register/                       # Methodology
 ├── ghm-status-sync/                       # Methodology
 ├── ghm-harvest/                           # Methodology
 └── ghm-sot-builder/                       # Methodology
 ```
 
-**29 skills total** covering the complete PRD lifecycle v0.1→v1.0.
+**44 skills total** covering the complete PRD lifecycle v0.1→v1.0, including framework-grounded v0.9 (Dunford / Hormozi / ORB) and v1.0 (Moore Chasm / Torres / Mom Test) skills.
 
 ---
 
@@ -77,8 +92,9 @@ skills/
 | **v0.5 Red Team**  | Risk Discovery Interview, Technical Stack Selection              | 2     |
 | **v0.6 Arch**      | Architecture Design, Technical Specification                     | 2     |
 | **v0.7 Build**     | Epic Scoping, Test Planning, Implementation Loop                 | 3     |
-| **v0.8 Release**   | Release Planning, Runbook Creation, Monitoring Setup             | 3     |
-| **v0.9 Launch**    | GTM Strategy, Launch Metrics, Feedback Loop Setup                | 3     |
+| **v0.8 Release**   | Release Planning, Runbook Creation, Monitoring Setup, Changelog-as-Marketing, Drift Baseline/Compare, Marketing-Ops Handoff | 6     |
+| **v0.9 Launch**    | GTM Strategy (orchestrator), Positioning (Dunford), Offer Construction (Hormozi), Launch Channels (ORB), AEO Audit, Alternatives Pages, Cold Outreach Tiered, HN/Reddit Launch, Launch Metrics, Feedback Loop Setup | 10    |
+| **v1.0 Adoption**  | Crossing the Chasm (Moore), Continuous Discovery (Torres), Mom Test Interview (Fitzpatrick), Case Study Builder, Testimonial Collector | 5     |
 | **Methodology**    | Gate Check, ID Register, Status Sync, Harvest, SoT Builder       | 5     |
 
 See [`skills-inventory.md`](skills-inventory.md) for full specifications.