@growthub/cli 0.7.0 → 0.7.2

package/assets/worker-kits/growthub-hyperframes-studio-v1/skills.md

@@ -0,0 +1,12 @@
+# Hyperframes Studio Skills
+
+1. Read `runtime-assumptions.md`.
+2. Read `docs/hyperframes-fork-integration.md`.
+3. Confirm target format: duration, ratio, channel.
+4. Produce:
+   - `templates/video-brief.md`
+   - `templates/composition-spec.md`
+   - `templates/render-plan.md`
+   - `templates/qa-checklist.md`
+
+Use deterministic command paths and avoid exposing secrets.

package/assets/worker-kits/growthub-hyperframes-studio-v1/templates/composition-spec.md

@@ -0,0 +1,6 @@
+# Composition Spec
+
+- Composition ID:
+- Timeline:
+- Scene breakdown:
+- Asset inputs:

package/assets/worker-kits/growthub-hyperframes-studio-v1/templates/qa-checklist.md

@@ -0,0 +1,6 @@
+# QA Checklist
+
+- [ ] Duration matches brief
+- [ ] Aspect ratio correct
+- [ ] Captions readable
+- [ ] Audio sync verified

package/assets/worker-kits/growthub-hyperframes-studio-v1/templates/render-plan.md

@@ -0,0 +1,13 @@
+# Render Plan
+
+## Commands
+
+```bash
+npx hyperframes preview
+npx hyperframes render
+```
+
+## Checks
+
+- Ensure deterministic assets
+- Record output path

package/assets/worker-kits/growthub-hyperframes-studio-v1/templates/video-brief.md

@@ -0,0 +1,7 @@
+# Video Brief
+
+- Goal:
+- Audience:
+- Duration:
+- Aspect ratio:
+- Delivery format:

package/assets/worker-kits/growthub-hyperframes-studio-v1/validation-checklist.md

@@ -0,0 +1,6 @@
+# Validation Checklist
+
+- [ ] `node setup/verify-env.mjs`
+- [ ] `bash setup/check-deps.sh`
+- [ ] `bash setup/clone-fork.sh`
+- [ ] All required output artifacts written

package/assets/worker-kits/growthub-hyperframes-studio-v1/workers/hyperframes-studio-operator/CLAUDE.md

@@ -0,0 +1,18 @@
+# Hyperframes Studio Operator
+
+**Kit:** `growthub-hyperframes-studio-v1`  
+**Worker ID:** `hyperframes-studio-operator`  
+**Version:** `1.0.0`
+
+## Role
+
+You convert campaign goals into implementation-ready Hyperframes composition and render artifacts.
+
+## Required startup
+
+1. Read `skills.md`.
+2. Read `runtime-assumptions.md`.
+3. Read `docs/hyperframes-fork-integration.md`.
+4. Verify environment with `node setup/verify-env.mjs`.
+
+If setup checks fail, stop and return remediation only.

package/assets/worker-kits/growthub-marketing-skills-v1/.env.example

@@ -0,0 +1,7 @@
+# Growthub Marketing Skills — Environment Config
+# Copy this file to .env and fill in keys as needed.
+# Never commit .env to version control.
+
+# Optional model/provider keys for enhanced research and generation.
+ANTHROPIC_API_KEY=your_anthropic_key_here
+OPENAI_API_KEY=your_openai_key_here

package/assets/worker-kits/growthub-marketing-skills-v1/QUICKSTART.md

@@ -0,0 +1,111 @@
+# Growthub Marketing Operator — Quickstart
+
+**Kit:** `growthub-marketing-skills-v1`
+**Worker:** `marketing-operator`
+**Upstream:** [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills)
+
+---
+
+## What This Kit Does
+
+The Growthub Marketing Operator is a self-contained AI agent environment for executing structured marketing workflows. It wraps proven frameworks from the open-source marketingskills library into a Growthub worker kit with brand context, templates, and output standards.
+
+Domains covered:
+- **CRO** — Landing page, signup flow, form, popup, paywall, and onboarding optimization
+- **SEO** — Technical audit, on-page, content quality, AI search visibility, schema markup
+- **Content** — Strategy, editorial planning, distribution, pillar mapping
+- **Email** — Welcome, nurture, re-engagement, cold outbound, event-based sequences
+- **Launch** — Pre-launch, launch day, post-launch phased playbooks
+- **Competitive** — Positioning gap analysis, feature matrices, messaging teardowns
+- **Growth** — Marketing ideas, pricing, referrals, free tools, churn prevention
+- **Copy** — Landing pages, ads, email, social — grounded in brand voice
+
+---
+
+## Setup — 4 Steps
+
+### Step 1 — Point Your Working Directory
+
+Export this kit to a local folder and point your AI agent's Working Directory at the kit root. All paths are relative to the kit root.
+
+### Step 2 — Copy the Environment File
+
+```bash
+cp .env.example .env
+```
+
+No API keys are required for core analysis. The operator works with any AI agent that can read markdown files.
+
+### Step 3 — Verify the Environment
+
+```bash
+node setup/verify-env.mjs
+```
+
+This confirms the kit structure is intact and all required files are present.
+
+### Step 4 — Set Up Your Product Context
+
+Copy the template and fill it in:
+
+```bash
+cp brands/_template/product-marketing-context.md brands/<your-company>/product-marketing-context.md
+```
+
+This is the single most important step. The product-marketing context grounds every deliverable in your specific product, audience, and positioning. Without it, the operator produces generic marketing advice instead of actionable recommendations.
+
+---
+
+## First Run
+
+1. Tell the operator: **"Run a CRO audit on [your landing page URL]"** or any marketing task
+2. The operator reads your product-marketing context for brand grounding
+3. The operator identifies the marketing domain and asks 3 clarifying questions
+4. The operator applies the appropriate framework and produces structured deliverables
+5. Output is saved to `output/<client-slug>/<project-slug>/`
+
+---
+
+## New Client Setup
+
+```bash
+cp brands/_template/product-marketing-context.md brands/<client-slug>/product-marketing-context.md
+# Fill in the 12 sections in the new file
+```
+
+The 12 sections: Product Overview, Target Audience, Personas, Problems & Pain Points, Competitive Landscape, Differentiation, Objections, Switching Dynamics, Customer Language, Brand Voice, Proof Points, Goals.
+
+---
+
+## Available Domains
+
+Tell the operator which marketing domain you need:
+
+| Domain | Example Request |
+|---|---|
+| CRO | "Audit this landing page for conversions" |
+| SEO | "Run an SEO audit on my site" |
+| Content | "Build a content strategy for my blog" |
+| Email | "Design a welcome email sequence" |
+| Launch | "Create a launch checklist for my new feature" |
+| Competitive | "Analyze my top 3 competitors" |
+| Copy | "Write homepage copy for my SaaS" |
+| Growth | "Give me 10 marketing ideas for my startup" |
+
+The operator will map your request to the correct framework and template automatically.
+
+---
+
+## Key Files
+
+| File | Purpose |
+|---|---|
+| `workers/marketing-operator/CLAUDE.md` | Agent operating instructions (start here) |
+| `skills.md` | Full methodology and skill dispatch table |
+| `brands/_template/product-marketing-context.md` | Blank product context template |
+| `brands/growthub/product-marketing-context.md` | Growthub reference example |
+| `templates/` | Frozen deliverable templates per domain |
+| `output/README.md` | Output directory structure and naming |
+| `docs/fork-integration.md` | How this kit relates to marketingskills |
+| `docs/skill-dispatch-methodology.md` | How skill routing works |
+| `docs/evaluation-frameworks.md` | Scoring and evaluation criteria |

package/assets/worker-kits/growthub-marketing-skills-v1/brands/_template/product-marketing-context.md

@@ -0,0 +1,147 @@
+# Product Marketing Context — [Company Name]
+
+> This file is the foundation for every marketing deliverable the operator produces. Fill in each section with real data — the more specific, the better the output. Use customer language, not polished marketing copy.
+
+---
+
+## 1. Product Overview
+
+| Field | Value |
+|---|---|
+| Product name | |
+| One-liner | |
+| Category | |
+| Product type | SaaS / Physical / Service / Marketplace / Other |
+| Pricing model | Free / Freemium / Subscription / One-time / Usage-based |
+| Starting price | |
+| Website | |
+
+---
+
+## 2. Target Audience
+
+| Field | Value |
+|---|---|
+| Company type | B2B / B2C / B2B2C |
+| Company size | |
+| Industry verticals | |
+| Decision-maker titles | |
+| Primary use case | |
+| Secondary use cases | |
+
+---
+
+## 3. Personas
+
+| Persona | Role | Key Motivation | Primary Objection |
+|---|---|---|---|
+| | | | |
+| | | | |
+
+---
+
+## 4. Problems & Pain Points
+
+1. **Core problem**: 
+2. **Cost of inaction**: 
+3. **Emotional tension**: 
+4. **Failed alternatives**: 
+
+---
+
+## 5. Competitive Landscape
+
+| Competitor | Type | Key Difference |
+|---|---|---|
+| | Direct | |
+| | Direct | |
+| | Indirect | |
+
+---
+
+## 6. Differentiation
+
+- **Key differentiator 1**: 
+- **Key differentiator 2**: 
+- **Key differentiator 3**: 
+- **Why customers choose you over alternatives**: 
+
+---
+
+## 7. Objections & Anti-Personas
+
+**Common objections:**
+1. 
+2. 
+3. 
+
+**Anti-personas** (who is NOT the right customer):
+- 
+
+---
+
+## 8. Switching Dynamics
+
+| Force | Description |
+|---|---|
+| Push (away from status quo) | |
+| Pull (toward your product) | |
+| Habit (keeps them where they are) | |
+| Anxiety (fear of switching) | |
+
+---
+
+## 9. Customer Language
+
+> Paste verbatim phrases from customer interviews, reviews, support tickets, or sales calls. These are more valuable than any polished positioning statement.
+
+- ""
+- ""
+- ""
+
+---
+
+## 10. Brand Voice
+
+| Attribute | Value |
+|---|---|
+| Tone | |
+| Style | |
+| Personality | |
+| Words we use | |
+| Words we avoid | |
+
+---
+
+## 11. Proof Points
+
+| Type | Detail |
+|---|---|
+| Key metric | |
+| Testimonial | |
+| Case study | |
+| Award/recognition | |
+
+---
+
+## 12. Goals
+
+| Goal | Metric | Target |
+|---|---|---|
+| Primary conversion action | | |
+| Secondary goal | | |
+| Revenue target | | |
+
+---
+
+## Messaging Guardrails
+
+- Never claim: 
+- Always include: 
+- Compliance notes: 
+
+---
+
+## Deliverables Log
+
+<!-- Append a line after every operator session -->

package/assets/worker-kits/growthub-marketing-skills-v1/brands/growthub/product-marketing-context.md

@@ -0,0 +1,145 @@
+# Product Marketing Context — Growthub
+
+---
+
+## 1. Product Overview
+
+| Field | Value |
+|---|---|
+| Product name | Growthub |
+| One-liner | Local-first AI agent runtime for go-to-market and developer experience teams |
+| Category | AI Development Tools / GTM Infrastructure |
+| Product type | SaaS + Open Source CLI |
+| Pricing model | Freemium (open-source CLI + hosted features) |
+| Starting price | Free (CLI), hosted plans TBD |
+| Website | growthub.ai |
+
+---
+
+## 2. Target Audience
+
+| Field | Value |
+|---|---|
+| Company type | B2B |
+| Company size | Startups to mid-market (5-200 employees) |
+| Industry verticals | SaaS, Developer Tools, Marketing Technology |
+| Decision-maker titles | Founders, Growth Leads, Marketing Directors, DevRel Leads |
+| Primary use case | Run AI agent worker kits locally for marketing and growth execution |
+| Secondary use cases | Content production pipelines, SEO auditing, email marketing automation |
+
+---
+
+## 3. Personas
+
+| Persona | Role | Key Motivation | Primary Objection |
+|---|---|---|---|
+| Solo Founder | CEO / Founder | Ship marketing without hiring an agency | "I don't have time to learn another tool" |
+| Growth Operator | Growth / Marketing Lead | Scale output without scaling headcount | "How is this different from ChatGPT?" |
+| Developer Advocate | DevRel / DX Engineer | Build agent-native workflows for the team | "Does it work with my existing stack?" |
+
+---
+
+## 4. Problems & Pain Points
+
+1. **Core problem**: Marketing execution doesn't scale — founders and small teams can't produce enough high-quality marketing output
+2. **Cost of inaction**: Slower growth, missed launch windows, inconsistent brand messaging across channels
+3. **Emotional tension**: Knowing what needs to be done but not having the bandwidth to do it well
+4. **Failed alternatives**: Generic AI chatbots produce disconnected, off-brand output that needs heavy editing
+
+---
+
+## 5. Competitive Landscape
+
+| Competitor | Type | Key Difference |
+|---|---|---|
+| ChatGPT / Claude (raw) | Indirect | No environment context, no brand grounding, no output standards |
+| Jasper / Copy.ai | Direct | Cloud-only, template-driven, no local execution or agent orchestration |
+| Generic prompt libraries | Indirect | No structure, no brand context, no quality bar enforcement |
+
+---
+
+## 6. Differentiation
+
+- **Key differentiator 1**: Local-first execution — your data stays on your machine
+- **Key differentiator 2**: Environment packaging — kits bundle prompts, templates, brand context, and output standards into one exportable unit
+- **Key differentiator 3**: Agent-agnostic — works with Claude Code, Cursor, Codex, Gemini, any local adapter
+- **Why customers choose you over alternatives**: Repeatable agent performance from structured environments, not scattered prompts
+
+---
+
+## 7. Objections & Anti-Personas
+
+**Common objections:**
+1. "I can just use ChatGPT for this" — You can, but without brand context and output standards, you'll spend more time editing than creating
+2. "This seems complex to set up" — One command to download a kit, point your agent at it, and start producing
+3. "What if I don't use Claude Code?" — Kits work with any agent that reads markdown — Cursor, Codex, Gemini, etc.
+
+**Anti-personas:**
+- Enterprise teams with existing full-service agencies and no interest in AI-assisted execution
+- Users who want a click-and-generate web app, not a local development environment
+
+---
+
+## 8. Switching Dynamics
+
+| Force | Description |
+|---|---|
+| Push | Frustration with generic AI output that doesn't match brand voice |
+| Pull | Structured environments that produce consistent, on-brand deliverables |
+| Habit | "I've been getting by with ChatGPT and manual editing" |
+| Anxiety | "Will this actually save time or just add another tool to learn?" |
+
+---
+
+## 9. Customer Language
+
+- "I need something that understands my product, not just generates generic marketing copy"
+- "We can't afford an agency but we need agency-quality output"
+- "I want to set it up once and have it produce consistent results every time"
+- "ChatGPT gives me a starting point but I spend 80% of the time editing"
+
+---
+
+## 10. Brand Voice
+
+| Attribute | Value |
+|---|---|
+| Tone | Direct, technical, confident without being arrogant |
+| Style | Concise — say more with fewer words |
+| Personality | Builder-first, practical, results-oriented |
+| Words we use | Ship, execute, build, operator, environment, local-first |
+| Words we avoid | Revolutionary, game-changing, disruptive, synergy, leverage |
+
+---
+
+## 11. Proof Points
+
+| Type | Detail |
+|---|---|
+| Key metric | 4 production worker kits shipped and publicly available |
+| Open source | CLI is MIT-licensed and published on npm |
+| Ecosystem | Works with 5+ AI agent platforms out of the box |
+
+---
+
+## 12. Goals
+
+| Goal | Metric | Target |
+|---|---|---|
+| Primary conversion action | CLI install (`npx @growthub/create-growthub-local`) | Growth in weekly installs |
+| Secondary goal | Kit downloads via discovery hub | Active kit exports per week |
+| Awareness | GitHub stars and contributor PRs | Community growth |
+
+---
+
+## Messaging Guardrails
+
+- Never claim specific revenue or ROI numbers without sourced data
+- Always position as "local-first" — never suggest data leaves the user's machine unless they opt into hosted features
+- Compliance notes: No testimonials that aren't from real users
+
+---
+
+## Deliverables Log
+
+<!-- Append a line after every operator session -->

package/assets/worker-kits/growthub-marketing-skills-v1/bundles/growthub-marketing-skills-v1.json

@@ -0,0 +1,43 @@
+{
+  "schemaVersion": 2,
+  "bundle": {
+    "id": "growthub-marketing-skills-v1",
+    "version": "1.0.0",
+    "kitId": "growthub-marketing-skills-v1",
+    "workerId": "marketing-operator"
+  },
+  "briefType": "marketing-operator",
+  "publicExampleBrandPaths": ["brands/growthub/product-marketing-context.md"],
+  "requiredFrozenAssets": [
+    "QUICKSTART.md",
+    ".env.example",
+    "skills.md",
+    "output-standards.md",
+    "runtime-assumptions.md",
+    "workers/marketing-operator/CLAUDE.md",
+    "brands/_template/product-marketing-context.md",
+    "brands/growthub/product-marketing-context.md",
+    "templates/cro-audit-brief.md",
+    "templates/seo-audit-report.md",
+    "templates/content-strategy-plan.md",
+    "templates/email-sequence-plan.md",
+    "templates/launch-checklist.md",
+    "templates/competitor-analysis.md",
+    "examples/landing-page-cro-sample.md",
+    "examples/seo-audit-sample.md",
+    "examples/email-sequence-sample.md",
+    "docs/fork-integration.md",
+    "docs/skill-dispatch-methodology.md",
+    "docs/evaluation-frameworks.md",
+    "setup/verify-env.mjs",
+    "output/README.md",
+    "growthub-meta/README.md",
+    "growthub-meta/kit-standard.md"
+  ],
+  "optionalPresets": [],
+  "export": {
+    "folderName": "growthub-agent-worker-kit-marketing-skills-v1",
+    "zipFileName": "growthub-agent-worker-kit-marketing-skills-v1.zip"
+  },
+  "activationModes": ["export"]
+}

package/assets/worker-kits/growthub-marketing-skills-v1/docs/evaluation-frameworks.md

@@ -0,0 +1,53 @@
+# Evaluation Frameworks
+
+## Frameworks by Domain
+
+### CRO — 7-Dimension Analysis
+
+Scoring: 0-3 per dimension, 0-21 composite.
+
+| Dimension | What It Measures |
+|---|---|
+| Value Proposition Clarity | Can users understand the offer within 5 seconds? |
+| Headline Effectiveness | Does headline match traffic source and promise a benefit? |
+| CTA Placement & Copy | Single primary action, benefit-driven, visible without scroll? |
+| Visual Hierarchy | Scannable layout, intentional whitespace, eye flow? |
+| Trust Signals | Logos, testimonials, reviews near decision points? |
+| Objection Handling | Price, fit, difficulty, risk addressed before CTA? |
+| Friction Points | Form fields, navigation, mobile experience? |
+
+### SEO — Priority Matrix
+
+| Priority | Criteria |
+|---|---|
+| Critical | Blocking indexation or causing ranking penalties |
+| High | Significant ranking impact, moderate fix effort |
+| Medium | Incremental improvement, low-moderate effort |
+| Low | Nice-to-have, minimal ranking impact |
+
+### SEO — E-E-A-T Assessment
+
+| Signal | What It Measures |
+|---|---|
+| Experience | First-hand experience signals in content |
+| Expertise | Depth, accuracy, and domain knowledge |
+| Authoritativeness | Source credibility and external recognition |
+| Trustworthiness | Accuracy, transparency, and safety signals |
+
+### Email — Sequence Quality Principles
+
+1. **One email, one job** — each message has a single purpose and CTA
+2. **Value before ask** — trust-building precedes conversion requests
+3. **Relevance over volume** — quality messaging outperforms frequency
+
+### Content — Gap Analysis Matrix
+
+| Axis | Categories |
+|---|---|
+| Funnel stage | Awareness / Consideration / Decision |
+| Content type | Blog / Guide / Tool / Case study / Comparison |
+| Competitor coverage | Covered / Gap / Opportunity |
+
+## Source Attribution
+
+These frameworks are adapted from [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills) (MIT license). The upstream library provides the detailed skill definitions; this document summarizes the evaluation criteria used by the operator.

package/assets/worker-kits/growthub-marketing-skills-v1/docs/fork-integration.md

@@ -0,0 +1,43 @@
+# Fork Integration — marketingskills
+
+## Upstream Repository
+
+- **Source**: [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills)
+- **License**: MIT
+- **Stars**: 21k+
+- **Skills**: 40+ marketing skill definitions
+
+## Relationship to This Kit
+
+This worker kit wraps the marketingskills library's frameworks and evaluation methodologies into a Growthub operator kit. The upstream library is a **reference**, not a runtime dependency.
+
+### What we took from the upstream:
+- Marketing domain taxonomy (CRO, SEO, Content, Email, Growth, etc.)
+- Evaluation frameworks (7-dimension CRO, E-E-A-T SEO, sequence architecture)
+- Cross-skill chaining patterns (which skills reference which)
+- Product-marketing-context as the foundational grounding document
+
+### What we added:
+- **Product-marketing context as brand kit** — the upstream's context file becomes our brand kit template with 12 structured sections
+- **Frozen deliverable templates** — structured output templates for each domain
+- **Operator workflow** — 7-step strict-order workflow with mandatory gates
+- **Output standards** — file naming, versioning, quality checklist
+- **Growthub examples** — reference deliverables using Growthub's own product context
+- **Kit packaging** — schemaVersion 2 manifest, bundle, activation modes
+
+### What we intentionally left out:
+- **Individual skill files** — skills are one dimension of the environment, not the centerpiece. The operator's skills.md consolidates all 40+ skills into a dispatch table and framework summaries. Users who want the full individual skill files can install them from the upstream repo alongside this kit.
+- **Tool integrations** — the upstream includes tool-specific integrations (GA4, Stripe, Mailchimp MCP). These are optional and not frozen into this kit. Users add them via their agent's tool configuration.
+- **Installation scripts** — the upstream's `npx skills add` flow is for standalone skill installation. This kit uses the Growthub kit export flow instead.
+
+## Updating from Upstream
+
+When the upstream library releases new skills or framework updates:
+
+1. Review the upstream changelog
+2. Update `skills.md` dispatch table if new skills are added
+3. Update framework summaries if evaluation criteria change
+4. Add new templates if new domains are introduced
+5. Bump the kit version in `kit.json`
+
+The upstream library evolves independently. This kit freezes a specific set of frameworks at a point in time. Not every upstream change needs to be pulled in.

package/assets/worker-kits/growthub-marketing-skills-v1/docs/skill-dispatch-methodology.md

@@ -0,0 +1,50 @@
+# Skill Dispatch Methodology
+
+## How the Operator Routes User Intent
+
+The marketing operator uses a two-phase routing system:
+
+### Phase 1 — Domain Detection
+
+The operator scans the user's request for trigger signals and maps to a primary marketing domain:
+
+| Signal Pattern | Domain |
+|---|---|
+| "conversions", "landing page", "bounce rate", "CRO", URL shared | CRO |
+| "rankings", "traffic", "SEO", "organic", "indexing" | SEO |
+| "content strategy", "blog", "editorial", "pillars" | Content |
+| "email", "sequence", "nurture", "cold outbound" | Email |
+| "launch", "go-to-market", "GTM", "product launch" | Launch |
+| "competitors", "alternatives", "positioning" | Competitive |
+| "write copy", "headline", "ad copy" | Copy |
+| "growth ideas", "referral", "pricing", "churn" | Growth |
+
+When signals are ambiguous or span multiple domains, the 3-question gate resolves the primary domain.
+
+### Phase 2 — Framework Selection
+
+Once the domain is identified, the operator loads the appropriate template and framework from `templates/`. Each framework defines:
+
+1. **Evaluation dimensions** — what to assess
+2. **Scoring criteria** — how to rate findings (where applicable)
+3. **Output structure** — required sections in the deliverable
+4. **Cross-references** — when to chain to another domain
+
+## Cross-Domain Chaining
+
+Skills are interconnected. The operator follows upstream-to-downstream ordering:
+
+```
+customer-research → content-strategy → copywriting
+competitor-analysis → pricing-strategy → launch-strategy
+seo-audit → content-strategy → programmatic-seo
+page-cro → copywriting → ab-test-setup
+```
+
+Each domain in the chain produces its own deliverable. Outputs are never merged across domains.
+
+## Relationship to Upstream Skills
+
+The dispatch table maps to 36+ skills from [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills). The operator consolidates these into 8 primary domains with framework summaries rather than loading individual skill files. This keeps the cognitive load manageable while preserving the evaluation rigor of the underlying frameworks.
+
+Users who want deeper skill-level granularity can install the upstream skills alongside this kit. The operator's dispatch table documents which upstream skill maps to which domain.