@growthub/cli 0.14.5 → 0.14.6

package/assets/worker-kits/growthub-email-marketing-v1/examples/nurture-sequence-sample.md

@@ -1,265 +0,0 @@
-# SAMPLE: Nurture Sequence — Growthub — Automation & AI Pillar — 5 Emails
-
-> Reference output. Shows a fully written 5-email nurture sequence.
-> Pillar: Automation & AI | Segment: new-lead | Persona: Performance director / CGO
-> Campaign brief: `examples/campaign-brief-sample.md`
-
----
-
-## EMAIL 1 OF 5 — THE REAL BOTTLENECK
-
-```
-EMAIL:    1 of 5
-THEME:    The real bottleneck
-SEND DAY: Day 0
-PILLAR:   Automation & AI
-SEGMENT:  new-lead
-```
-
-**Subject:** The thing most performance teams miss about scaling creative  
-**Preview:** It's not the budget. It's not the team. It's this.
-
----
-
-Here's something we see across almost every performance team we work with:
-
-The budget is there. The team is capable. The strategy is solid. And yet creative is still the constraint.
-
-Not because they're not producing enough. Because they're producing on a cycle that's too slow to keep up with what the market is telling them.
-
-The average creative production cycle runs 2–4 weeks. From insight to live ad. By the time a concept is in market, the data that inspired it is already three weeks old.
-
-The bottleneck isn't volume. It's the gap between having a good idea and learning whether it works.
-
-We built a framework around closing that gap. You can download it here:
-
-**Download the Creative Intelligence Framework →** [URL]
-
----
-
-**PS:** This framework is the same one we use to brief every campaign we produce. Take it and apply it — if you want to see how we'd use it on your account specifically, reply and we'll go from there.
-
----
-
-**PLAIN-TEXT FALLBACK:**
-```
-Subject: The thing most performance teams miss about scaling creative
-Preview: It's not the budget. It's not the team. It's this.
-
-Here's something we see across almost every performance team we work with:
-
-The budget is there. The team is capable. The strategy is solid. And yet creative is still the constraint.
-
-Not because they're not producing enough. Because they're producing on a cycle that's too slow to keep up with what the market is telling them.
-
-The average creative production cycle runs 2-4 weeks. From insight to live ad. By the time a concept is in market, the data that inspired it is already three weeks old.
-
-The bottleneck isn't volume. It's the gap between having a good idea and learning whether it works.
-
-We built a framework around closing that gap. You can download it here: [URL]
-
-P.S. This framework is the same one we use to brief every campaign we produce. Take it and apply it - if you want to see how we'd use it on your account specifically, reply and we'll go from there.
-
----
-You're receiving this because you opted in at [source]. Unsubscribe: [link] | Growthub, [address]
-```
-
----
-
-## EMAIL 2 OF 5 — THE INSIGHT
-
-```
-EMAIL:    2 of 5
-THEME:    The insight
-SEND DAY: Day +3
-PILLAR:   Automation & AI
-SEGMENT:  nurture-active
-```
-
-**Subject:** Why your creative cycle is slower than it looks  
-**Preview:** The delay isn't in production. It's in the step before it.
-
----
-
-Most performance teams think their production cycle is their slowest step.
-
-It's not. The slowest step is the one before production starts.
-
-The gap between "we know this hook type is working" and "we have a brief ready for the next round" is usually 5–10 days. Not because anyone is slow — because the process for extracting learning and turning it into a new brief doesn't exist in a structured way.
-
-What you end up with is a production cycle that starts from scratch every time. No compounding. No pattern library. Just a new brief that borrows informally from what worked before.
-
-The fix is simpler than it sounds: build an intelligence layer between your testing and your briefing. A process — not a tool — that extracts the "why it worked" from every ad and feeds it directly into the next brief.
-
-That's the one change that separates programs that compound from programs that plateau.
-
-Reply and tell me — is this where your cycle is leaking, or is the constraint somewhere else in the process?
-
----
-
-**PS:** Next email, I'll walk through the specific 3-step framework we use to build this layer. Takes about 5 minutes to read.
-
----
-
-**PLAIN-TEXT FALLBACK:**
-```
-Subject: Why your creative cycle is slower than it looks
-Preview: The delay isn't in production. It's in the step before it.
-
-Most performance teams think their production cycle is their slowest step.
-
-It's not. The slowest step is the one before production starts.
-
-The gap between "we know this hook type is working" and "we have a brief ready for the next round" is usually 5-10 days. Not because anyone is slow - because the process for extracting learning and turning it into a new brief doesn't exist in a structured way.
-
-What you end up with is a production cycle that starts from scratch every time. No compounding. No pattern library. Just a new brief that borrows informally from what worked before.
-
-The fix is simpler than it sounds: build an intelligence layer between your testing and your briefing. A process - not a tool - that extracts the "why it worked" from every ad and feeds it directly into the next brief.
-
-That's the one change that separates programs that compound from programs that plateau.
-
-Reply and tell me - is this where your cycle is leaking, or is the constraint somewhere else?
-
-P.S. Next email, I'll walk through the specific 3-step framework we use to build this layer. Takes about 5 minutes to read.
-
----
-You're receiving this because [source]. Unsubscribe: [link] | Growthub, [address]
-```
-
----
-
-## EMAIL 3 OF 5 — THE FRAMEWORK
-
-```
-EMAIL:    3 of 5
-THEME:    The framework
-SEND DAY: Day +6
-PILLAR:   Automation & AI / Education & Strategy
-SEGMENT:  nurture-active
-```
-
-**Subject:** The 3-step creative intelligence framework  
-**Preview:** This is the system. Steal it.
-
----
-
-Last email I mentioned the intelligence layer — the missing step between testing and briefing.
-
-Here's how it actually works:
-
-**1. Production layer** — the ability to generate and test concepts fast. AI UGC or lean UGC production. Goal: collapse the cycle from 3 weeks to 3 days.
-
-**2. Intelligence layer** — the process of extracting *why* a concept won. Hook analysis, pain-point mapping, pattern logging. Goal: every test adds to a library, not just to a report.
-
-**3. Iteration layer** — the system that uses the intelligence layer to brief the next round faster. Goal: each cycle is smarter than the last. Cost-per-learning drops. Win rate compounds.
-
-Most programs have layer 1. Very few have layer 2. Almost none have all three connected.
-
-When all three are in place, you stop testing randomly and start building a library of what works — specific to your audience, your product, and your market.
-
-You can take this framework and start building it yourself. Or, if you want to see how we'd implement it for your account specifically:
-
-**Download the full framework → [URL]**
-
----
-
-**PS:** Next email, I'll show you what this looked like for a specific brand — the before, the after, and the numbers.
-
----
-
-## EMAIL 4 OF 5 — THE PROOF
-
-```
-EMAIL:    4 of 5
-THEME:    The proof
-SEND DAY: Day +10
-PILLAR:   Client Results
-SEGMENT:  nurture-engaged
-```
-
-**Subject:** How one team went from 2-week cycles to weekly wins  
-**Preview:** Same budget. Same team. One structural change.
-
----
-
-A performance director at a DTC brand came to us spending $150K/month on Meta. Budget to scale. Creative team already stretched.
-
-They were running on a 3-week production cycle. Brief to live ad: 21 days. Testing maybe 4–6 new concepts per month.
-
-The constraint wasn't the creative team. It was the process. Every brief started from scratch. There was no systematic way to extract learning from winning ads and feed it back into the next round.
-
-We restructured their production pipeline around the 3-layer framework. Production layer first — collapsed the cycle to under a week using AI-assisted UGC production. Intelligence layer second — built a pattern log that captured hook type, proof structure, and audience signal from every ad. Iteration layer third — briefs now started from the pattern log, not from scratch.
-
-Within 60 days: 5x more concepts tested per month. Cost-per-test down 60%. Win rate up because every brief was starting from learning, not a blank page.
-
-Same team. Same budget. One structural change.
-
-If you're spending $50K or more on ads and creative is the constraint — this is the conversation worth having.
-
-**Book your growth call → [URL]**
-
----
-
-**PS:** If you're not ready for a call yet, reply to this email and tell me where you are in the process. I'll send you the resource that's most relevant to your specific situation.
-
----
-
-## EMAIL 5 OF 5 — THE ACTIVATION
-
-```
-EMAIL:    5 of 5
-THEME:    The activation
-SEND DAY: Day +14
-PILLAR:   Automation & AI
-SEGMENT:  nurture-activation
-```
-
-**Subject:** Still hitting the creative ceiling? Here's the thing.  
-**Preview:** Most of the work is already done. This is just the next step.
-
----
-
-Over the past two weeks, we've covered:
-
-- Why the production cycle gap is the real constraint (not volume, not budget)
-- The intelligence layer — the missing step between testing and learning
-- The 3-layer framework that connects production, intelligence, and iteration
-- What it looked like for one team that rebuilt their creative pipeline
-
-If any of that landed — if the creative ceiling is still the constraint on your ability to scale — the next step is straightforward.
-
-20 minutes. We look at your current setup, identify exactly where the gap is, and show you what a restructured pipeline would look like for your account.
-
-No pitch. No pressure. Worst case: you leave with a clear picture of where the bottleneck actually is.
-
-**Book your growth call → [URL]**
-
----
-
-**PS:** This is the last email in this sequence. If the timing isn't right, no problem — hold onto this. When it is, reply to this email and I'll pick it up from here.
-
----
-
-**PLAIN-TEXT FALLBACK:**
-```
-Subject: Still hitting the creative ceiling? Here's the thing.
-Preview: Most of the work is already done. This is just the next step.
-
-Over the past two weeks, we've covered:
-
-- Why the production cycle gap is the real constraint (not volume, not budget)
-- The intelligence layer - the missing step between testing and learning
-- The 3-layer framework that connects production, intelligence, and iteration
-- What it looked like for one team that rebuilt their creative pipeline
-
-If any of that landed - the next step is straightforward.
-
-20 minutes. We look at your current setup, identify where the gap is, and show you what a restructured pipeline would look like for your account.
-
-Book your growth call: [URL]
-
-P.S. This is the last email in this sequence. If the timing isn't right, no problem - hold onto this. When it is, reply to this email and I'll pick it up from here.
-
----
-Unsubscribe: [link] | Growthub, [address]
-```

package/assets/worker-kits/growthub-email-marketing-v1/examples/subject-line-matrix-sample.md

@@ -1,116 +0,0 @@
-# SAMPLE: Subject Line Matrix — Growthub Nurture Sequence — Email 1 of 5
-
-> Reference output. Shows a fully completed subject line matrix for Email 1 of the nurture sequence.
-> Campaign: Creative Velocity Nurture — Q2 2026
-> Use as a quality benchmark for subject line matrix outputs.
-
----
-
-## MATRIX METADATA
-
-```
-CAMPAIGN:     Growthub Creative Velocity Nurture — Q2 2026
-CLIENT:       growthub
-PILLAR:       Automation & AI
-SEGMENT:      new-lead
-VERSION:      v1
-DATE:         2026-04-10
-```
-
----
-
-## EMAIL 1 — THE REAL BOTTLENECK
-
-> Theme: The real bottleneck | Send Day: Day 0 | Pillar: Automation & AI
-
-| # | Type | Subject Line | Preview Text | Notes |
-|---|---|---|---|---|
-| 1A | Curiosity | The thing most performance teams miss about scaling creative | It's not the budget. It's not the team. It's this. | Strong loop — doesn't reveal the answer |
-| 1B | Curiosity | Your creative bottleneck isn't where you think it is | The constraint is 2 steps upstream from where you're looking. | Reframe — challenges their current diagnosis |
-| 1C | Social Proof | What brands scaling past $300K/month on Meta have in common | It's not what you'd expect — and it's fixable. | Proof-forward — uses aspirational metric as anchor |
-| 1D | Personal | [First Name] — quick question about your creative pipeline | Is production speed the constraint, or is it something else? | Works best for smaller, targeted lists |
-| 1E | Direct | Scale ad spend without increasing creative production costs | Here's the system behind it. | Direct value prop — good for warm-ish new leads |
-
-**Recommended A/B test:** `1A` vs `1B`  
-**Rationale:** Both are curiosity-type but with different reframe angles. 1A is broader (works for any performance team). 1B is a more provocative reframe (works better for leads who are already frustrated and have already tried to diagnose the problem). Testing these two will reveal which diagnostic frame resonates with this segment.
-
----
-
-## EMAIL 2 — THE INSIGHT
-
-> Theme: The insight | Send Day: Day +3 | Pillar: Automation & AI
-
-| # | Type | Subject Line | Preview Text | Notes |
-|---|---|---|---|---|
-| 2A | Curiosity | Why your creative cycle is slower than it looks | The delay isn't in production. It's in the step before it. | Names the hidden constraint |
-| 2B | Curiosity | Nobody talks about this part of the testing process | It's the step between running ads and knowing why they worked. | Pattern: "nobody talks about" — works for education angle |
-| 2C | Direct | The missing step between creative testing and scaling | Here's what it is and why it matters. | Clear value prop — for direct preference audiences |
-| 2D | Social Proof | What high-win-rate creative programs have in common | It's not the production quality. It's this structural difference. | Curiosity + proof hybrid |
-| 2E | Urgency | Every week your testing cycle doesn't compound, your competitors' does | The gap is growing. Here's how to close it. | Stronger urgency angle — use if 1A or 1B underperformed in Email 1 |
-
-**Recommended A/B test:** `2A` vs `2D`  
-**Rationale:** If 1A won Email 1, 2A is the logical follow-through (same curiosity tone). 2D adds a proof element for those who responded to the social proof angle. Testing these reveals whether this audience is more insight-driven or proof-driven.
-
----
-
-## EMAIL 3 — THE FRAMEWORK
-
-> Theme: The framework | Send Day: Day +6 | Pillar: Automation & AI / Education
-
-| # | Type | Subject Line | Preview Text | Notes |
-|---|---|---|---|---|
-| 3A | Curiosity | The 3-step creative intelligence framework | This is the system. Steal it. | Direct education — strong for education-pillar campaigns |
-| 3B | Direct | How to test 5x more concepts without increasing production cost | The framework behind it — and how to build it yourself. | Outcome-first — works for metric-focused readers |
-| 3C | Social Proof | The framework behind every high-win-rate creative program | Most brands have layer 1. Very few have all three. | Proof via pattern — "here's what winners do" |
-| 3D | Curiosity | Most creative programs don't compound. Here's why. | And the one structural change that fixes it. | Problem reframe — strong for frustrated prospects |
-| 3E | Personal | [First Name] — the framework I mentioned | 3-step, 5 minutes to read. Here it is. | Works if 1D/personal performed well in earlier emails |
-
-**Recommended A/B test:** `3A` vs `3C`  
-**Rationale:** 3A is pure education (appeals to readers who engaged with the framework in Emails 1–2). 3C positions the same framework as something that proven winners already use (proof validation). Which one wins tells you whether this segment is more motivated by learning or by proof of results.
-
----
-
-## EMAIL 4 — THE PROOF
-
-> Theme: The proof | Send Day: Day +10 | Pillar: Client Results
-
-| # | Type | Subject Line | Preview Text | Notes |
-|---|---|---|---|---|
-| 4A | Social Proof | How one team went from 2-week cycles to weekly wins | Same budget. Same team. One structural change. | Strong proof story — specific and believable |
-| 4B | Social Proof | 5x more creative tests per month — the system behind it | No new hire. No budget increase. Here's what changed. | Metric-led — works for data-first readers |
-| 4C | Curiosity | What changed for a $150K/month brand in 60 days | It wasn't what they expected — but it worked. | Curiosity + proof hybrid |
-| 4D | Urgency | The creative ceiling doesn't lift itself | Here's the proof that the right system does. | Urgency into proof — works for late-stage nurture |
-| 4E | Direct | See the result: creative pipeline rebuilt in 60 days | One team. One structural change. Here's exactly what happened. | Direct — works for warm, engaged readers |
-
-**Recommended A/B test:** `4A` vs `4B`  
-**Rationale:** 4A leads with the story (narrative readers). 4B leads with the metric (data-first readers). This is the most important A/B test in the sequence — the winner reveals whether this segment makes decisions from stories or from numbers.
-
----
-
-## EMAIL 5 — THE ACTIVATION
-
-> Theme: The activation | Send Day: Day +14 | Pillar: Automation & AI
-
-| # | Type | Subject Line | Preview Text | Notes |
-|---|---|---|---|---|
-| 5A | Urgency | Still hitting the creative ceiling? Here's the thing. | Most of the work is already done. This is just the next step. | Activation urgency — references the journey |
-| 5B | Direct | Your growth call is ready to book | 20 minutes. We look at your setup. You leave with a clear picture. | Straightforward — works for direct preference readers |
-| 5C | Urgency | Last email — and the thing worth doing before I go quiet | One step that might change how you're thinking about this. | Last email tension — honest and clean |
-| 5D | Curiosity | The question that changes this conversation | If you can answer yes to this, the rest is easy. | Curiosity into activation — unusual but effective for engaged lists |
-| 5E | Personal | [First Name] — last one from me on this | If the timing's ever right, here's the door. | Warm close — works for re-engagement into the sequence |
-
-**Recommended A/B test:** `5A` vs `5C`  
-**Rationale:** 5A references the sequence journey (works for engaged readers who have been through all 5 emails). 5C uses the "last email" tension honestly (works for readers who haven't engaged much and need a reason to act now).
-
----
-
-## SUBJECT LINE RULES APPLIED IN THIS MATRIX
-
-- All subject lines under 50 characters (mobile-safe)
-- No spam trigger words used
-- No ALL CAPS used
-- Preview text extends the subject — never repeats it
-- All [First Name] variants included only in personal-type rows
-- Urgency language used only in Emails 4 and 5 — not in early nurture
-- A/B test rationale provided for each email
-- Every variant is genuinely different — not just word swaps

package/assets/worker-kits/growthub-email-marketing-v1/growthub-meta/README.md

@@ -1,108 +0,0 @@
-# Growthub Agent Worker Kit — Email Marketing Strategist v1
-
-**Kit ID:** `growthub-email-marketing-v1`  
-**Version:** `1.0.0`  
-**Type:** `worker`  
-**Execution mode:** `export`  
-**Schema version:** `2`
-
----
-
-## What this kit does
-
-This kit gives an agent a complete, self-contained environment for Growthub-focused email marketing work. When exported and pointed at via the agent Working Directory, the agent can:
-
-- Write brand-aware email copy aligned to Growthub's voice, positioning, and messaging pillars
-- Plan campaigns around Growthub's 5 content pillars
-- Structure nurture, outbound, follow-up, promotional, and re-engagement sequences
-- Produce implementation-ready output artifacts (campaign briefs, sequence plans, email drafts, subject line matrices, CTA matrices, QA checklists)
-- Push final draft outputs to an email platform (GHL or any future platform integration) via the platform adapter pattern documented in `runtime-assumptions.md`
-
----
-
-## How to activate
-
-1. Export the kit: `growthub kit download growthub-email-marketing-v1`
-2. Point the agent's Working Directory at the exported folder
-3. The agent reads `workers/email-marketing-strategist/CLAUDE.md` as its entrypoint
-4. `skills.md` is the source of truth for methodology — the agent reads it first on every session
-
----
-
-## Folder structure
-
-```
-growthub-email-marketing-v1/
-├── kit.json                                     ← master manifest
-├── bundles/
-│   └── growthub-email-marketing-v1.json         ← export bundle definition
-├── growthub-meta/
-│   ├── README.md                                ← this file
-│   └── kit-standard.md                          ← locked contributor contract
-├── skills.md                                    ← master methodology
-├── output-standards.md                          ← output format contract
-├── runtime-assumptions.md                       ← platform integration docs
-├── workers/
-│   └── email-marketing-strategist/
-│       └── CLAUDE.md                            ← agent entrypoint
-├── brands/
-│   ├── _template/brand-kit.md                   ← blank template
-│   └── growthub/brand-kit.md                    ← Growthub canonical brand kit
-├── templates/
-│   ├── campaign-brief-template.md
-│   ├── sequence-planner.md
-│   ├── email-draft.md
-│   ├── subject-line-matrix.md
-│   ├── cta-matrix.md
-│   ├── qa-checklist.md
-│   ├── email-formats/                           ← 5 frozen campaign formats
-│   ├── email-modules/                           ← modular copy primitives
-│   └── hooks-library/                           ← subject line pattern library
-└── examples/                                    ← reference outputs
-```
-
----
-
-## Supported output types
-
-| Output | Description |
-|---|---|
-| `campaign-brief` | Full campaign spec — pillar, segment, format, arc, CTA, compliance |
-| `email-sequence` | Multi-email sequence plan + individual email drafts |
-| `broadcast` | Single one-off email draft |
-| `subject-line-matrix` | 5+ subject line variants per email with preview text |
-| `cta-matrix` | Primary, soft, and reply CTA variants |
-| `qa-checklist` | Pre-send review checklist |
-
----
-
-## Email platform integration
-
-This kit uses a platform-agnostic adapter pattern. GoHighLevel (GHL) is the reference implementation. See `runtime-assumptions.md` for the full interface spec, GHL adapter details, and how to extend to additional platforms.
-
----
-
-## Setup notes
-
-- No build step required — all outputs are Markdown
-- No npm dependencies — the kit runs with the agent's native tools
-- Browser use is optional but supported (see `runtime-assumptions.md`)
-- API credentials must be set as environment variables — never stored in kit files
-
----
-
-## File naming convention
-
-```
-output/<client-slug>/<campaign-slug>/<ClientSlug>_<CampaignType>_<Pillar>_v<N>_<YYYYMMDD>.md
-```
-
-Example: `output/growthub/q2-nurture/Growthub_Nurture_Automation_v1_20260410.md`
-
----
-
-## Version history
-
-| Version | Date | Notes |
-|---|---|---|
-| 1.0.0 | 2026-04-10 | Initial release |

package/assets/worker-kits/growthub-email-marketing-v1/growthub-meta/kit-standard.md

@@ -1,87 +0,0 @@
-# Kit Standard — Email Marketing Strategist v1
-
-This document defines the locked contract for this kit. It governs what may and may not change across versions, and what every contributor and agent must respect.
-
----
-
-## Required files contract
-
-Every valid release of this kit MUST contain all paths listed in `kit.json → frozenAssetPaths`. No path may be removed without a version bump and schema review.
-
-Files are frozen. They may be updated only through a versioned release process — not in-place hotfixes.
-
----
-
-## Bundle contract
-
-The bundle manifest at `bundles/growthub-email-marketing-v1.json` MUST remain consistent with `kit.json`:
-
-- `bundle.kitId` must match `kit.id`
-- `bundle.workerId` must match `entrypoint.workerId`
-- `requiredFrozenAssets` must be a subset of `frozenAssetPaths`
-- `export.folderName` and `export.zipFileName` must not change within a minor version
-
----
-
-## Export boundary
-
-When exported via `growthub kit download`, this kit produces a self-contained folder. The export MUST be complete and independently operational — the agent should require no external files to function.
-
-Sensitive credentials (API keys, account IDs) are NEVER stored in kit files. They are always injected via environment variables at runtime. See `runtime-assumptions.md`.
-
----
-
-## Brand kit rules
-
-- `brands/_template/brand-kit.md` is the canonical template. It must contain every field an agent needs to operate.
-- `brands/growthub/brand-kit.md` is the reference example. It must be fully populated.
-- Client-specific brand kits added by agents or operators at runtime are NOT part of the frozen asset set and are NOT shipped in the export.
-- The `publicExampleBrandPaths` in `kit.json` and the bundle are the only brand files shipped in the export. Do not add client-specific files to `frozenAssetPaths`.
-
----
-
-## Agent operating rules
-
-The agent entrypoint at `workers/email-marketing-strategist/CLAUDE.md` is the law. It governs:
-
-- Step order (steps may not be skipped)
-- The 3-question gate before any writing begins
-- Output structure and file naming
-- Platform integration mode (API vs browser-assisted)
-- Deliverables logging
-
-`skills.md` is the methodology. It is the source of truth for HOW the agent works. `CLAUDE.md` references `skills.md` — agents must read it before beginning any task.
-
----
-
-## Output rules
-
-- All outputs are Markdown files
-- Outputs go to `output/<client-slug>/<campaign-slug>/` relative to the working directory
-- Files are versioned (`v1`, `v2`, ...) — never overwrite an existing version
-- Deliverables log must be updated in the client brand kit after every completed output
-- Plain-text fallback blocks are required in every email draft
-
----
-
-## Platform integration rules
-
-- Email platform integration follows the adapter pattern defined in `runtime-assumptions.md`
-- GHL is the reference adapter
-- No platform-specific logic is hardcoded into `skills.md`, `CLAUDE.md`, or any template file
-- New platform adapters are added as new files — they do not modify existing kit files
-
----
-
-## What is allowed to change between versions
-
-| Change type | Allowed without major version bump |
-|---|---|
-| Adding new email format files | Yes (minor) |
-| Adding new module files | Yes (minor) |
-| Adding rows to subject-line-patterns.csv | Yes (patch) |
-| Updating brand kit guidance text | Yes (minor) |
-| Changing frozenAssetPaths | No — requires minor version bump |
-| Changing bundle export names | No — requires major version bump |
-| Changing worker entrypoint path | No — requires major version bump |
-| Adding a new platform adapter | Yes (minor, in new file only) |

package/assets/worker-kits/growthub-email-marketing-v1/helpers/README.md

@@ -1,29 +0,0 @@
-# `helpers/` — safe shell tool layer (primitive #6)
-
-A **helper** is a small, deterministic script an agent invokes via one shell call instead of reconstructing a raw pipeline inline. Helpers are reviewable, deterministic, and safer than raw commands.
-
-## Convention
-
-```
-helpers/
-├── README.md                 # this file
-├── <verb>.sh                 # single-purpose shell scripts
-├── <verb>.mjs                # optional: Node-based helper
-└── <verb>.py                 # optional: Python-based helper
-```
-
-Every helper:
-
-1. Carries a one-line header comment: what it does + how to invoke it.
-2. Uses `set -euo pipefail` (or equivalent) — no silent failures.
-3. Has a matching row in the parent `SKILL.md`'s `helpers[]` frontmatter.
-4. Is referenced from `skills.md` at the step the agent should invoke it (no duplicated shell bodies).
-
-## When to promote an inline snippet
-
-- Appears more than once in `skills.md`.
-- Has fragile quoting or path interpolation.
-- Calls a side-effecting binary (ffmpeg, git, gh, npm install).
-- Multiple sub-skills invoke it.
-
-See the parent `SKILL.md` `helpers[]` array for the current roster.