/marketing-os-newsletter

name: marketing-os-newsletter
description: "Write a newsletter edition from the Marketing OS, gated step by step, then file it as a real asset. Reads the ICP, voice register, offer and live price, measured patterns and swipe file from the OS, never from bundled copies. Sources from a pillar asset, the values doc, a community post or a raw idea. Fans out five parallel readers over the OS before the first gate, and verifies the draft with independent agents against the voice register and offer facts. Seven gates with options at each; the edition is written only after the angle, insight, outcome, outline, subject line and hook are confirmed. On approval it writes Channels/newsletter/broadcasts/YYYY-MM-DD-slug.md, appends the shipped subject line to the swipe file, updates the parent's repurpose tree, logs, and always closes by rendering the edition. Use when the user says 'write a newsletter', or runs /marketing-os-newsletter.\\\"\ndisable-model-invocation: true\nargument-hint: \\\"[source, e.g. a published video slug, 'values doc', or a raw idea]"
disable-model-invocation: true

Marketing OS Newsletter

Write one edition, in the operator's voice, from what the OS already knows. Then put it back in the OS.

**This skill carries no context of its own.** No ICP file, no offer summary, no voice guide, no example folder. Every one of those exists once in the OS and this skill reads it there. That is the whole point: the original newsletter skill shipped ten reference files and six of them were copies that went stale the moment the business changed. A price baked into a skill is wrong within days.

Run from the OS root. **Stay inside that root.** Read and write only within it, and do not go looking for other vaults on the machine. **Start from zero on identity.** Every name, handle, price and number comes from the OS or from the user in this conversation, never from your context, the system username or the folder name. If the OS does not name it, you do not know it.

First, check the OS is there

| State | Do | | --- | --- | | No `Context/config.md` | Not a Marketing OS. Point at `marketing-os-setup` and stop | | No `Channels/newsletter/` | The channel does not exist here. Say so and stop | | Both exist, `broadcasts/` empty | Run anyway. Say the style ground truth is missing and name what would fix it | | Both exist, `broadcasts/` populated | Normal run |

The read map

Read these before drafting, and read them from the OS every run. Full contract in `references/os-contract.md`.

| What you need | Where it lives | | --- | --- | | Instance literals: handles, cadence, connectors | `Context/config.md` | | Who the reader is, and the pain to name | `Context/icp/*.md`, one file per segment | | The master voice register | `Context/personal-brand/voice.md` | | The newsletter delta only | `Channels/newsletter/voice.md` | | Role in the funnel, format contract, cadence | `Channels/newsletter/strategy.md` | | Subject line mechanics | `Channels/newsletter/sop-subject-line-playbook.md` | | The edition structure | `Intelligence/research/frameworks/newsletter-structure.md` | | Proven openers, subjects, CTAs | `Intelligence/research/swipe/` | | **The style ground truth** | `Channels/newsletter/broadcasts/`, the real sends | | What has actually performed | `Analytics/what-works.md` | | Verbatim customer language | `Intelligence/research/voice-of-customer.md` | | The promise, the proof, **the live price** | `Offers/{offer}/offer.md` and `Offers/{offer}/proof/` | | Operator background, only if the edition needs it | `Context/personal-brand/background.md` |

**Precedence when they disagree.** The real sends in `broadcasts/` beat the structure framework, which beats the swipe file, which beats the voice register. A living example outranks a rule written about it.

**Read only what the edition needs.** A repurpose of a published video does not need the background file. Do not load the whole OS.

The four properties this skill exists to hold

Every gate below is one of these. They are the reason a gated process beats a one-shot draft, and cutting any of them collapses the quality.

1. **The insight comes from the operator, not from you.** If you generate the substance, the edition is generic and anyone could send it. You supply structure, options and craft. 2. **Human in the loop, one gate at a time.** Never write the edition before the gates are cleared. 3. **Options at every gate.** Three to five real choices beat one draft to correct. Correcting is slow, choosing is fast. 4. **Ground it in real examples.** The `broadcasts/` folder is the ground truth and it grows every time this skill runs.

Before Gate 0: fan out to load the OS

**Launch these readers in one batch, before you ask the operator anything.** Reading the OS serially, or skimming three files and starting the gates, is what produces generic options at Gate 0 and a draft that could have come from any skill. The gates are only as good as what is behind them.

| Agent | Reads | Returns | | --- | --- | --- | | 1 | `Channels/{email channel}/broadcasts/`, the six most recent sends **in full** | The mechanics actually in use: paragraph length, the turn, the sign-off, the CTA shapes, and **which material is already spent** | | 2 | All of `Context/icp/*.md` | Every segment, its pain in the reader's own words, and which are content audiences versus buyer segments | | 3 | `Intelligence/research/swipe/` and `Channels/{email channel}/sop-subject-line-playbook.md` | The proven openers, subject lines and CTAs, and the rules governing them | | 4 | `Analytics/what-works.md` and `Analytics/metrics.md` | The measured findings and what each licenses, the untested beliefs labelled as such, and every number with its pull date | | 5 | `Offers/{offer}/offer.md` and `Offers/{offer}/proof/` | The promise, **the live price table with effective dates**, and the real proof with sources, or the fact that the folder is empty |

Plus, once a source is chosen, a six