Os Message Architecture

• Source: Built by Jonathan Kvarfordt. Standalone bundle v0.2.
• Frameworks applied: Act

It covers a job one of the eight revenue seats has to finish, and it states its process, decision rules, and output format clearly enough for an agent to follow.

1. 1.Copy the SKILL.md text below, or download the raw file.
2. 2.Create a folder named exactly os-message-architecture in your agent's skills directory.
3. 3.Save the file inside that folder as SKILL.md.
4. 4.Ask the agent one of the trigger requests below.
5. 5.Check the output against what you already know before it leaves your desk.

• write the outreach
• what do I say to this account
• draft the email for X

---
name: os-message-architecture
description: Writes outreach for one account from its Account Context using the path its evidence allows (confirmed pain, signal-only question, or research task), mapped to three decision roles, with every proof point drawn from the proof register. Use when the user says "write the outreach", "what do I say to this account", "draft the email for X", "message architecture", or as the ACT step after scoring.
---

# os-message-architecture

Produce `artifacts/os-message-<account-slug>.md`. The evidence decides the path. The path decides what the message is allowed to claim.

## Preflight (every run)
1. Confirm `GTM-AI-OS/` is accessible. If not: "I need access to your GTM-AI-OS folder. Select it in Cowork's folder picker, or run `os-install` from the core plugin." Stop until it is.
2. Read `CLAUDE.md`, `INDEX.md`, `context/fence.md`, and the last 20 lines of `signals/log.md`.
3. Check the prerequisites below. For anything missing, ask the cold-start questions and offer: (a) run the earlier skill now, or (b) fast path: answer the questions and this skill writes a minimal version of the missing artifact marked `status: draft (fast path)`.
4. Before any connector write or external send, check `context/fence.md` for the account's bucket and the action. Auto = proceed. Approve = ask. Never = decline and name the row.
5. On finish, update the artifact row in `INDEX.md` (with `[[wikilinks]]` to the files this run read) and append one line to `signals/log.md`: `<date> | <skill> | <files read> | <produced> | <human changed> | <open question>`.

## Prerequisites
- A full Account Context record for the account, scored by `os-account-scorer`.
- `context/messaging.md` with a thesis, tone rules, and a proof register.
Cold-start: "Give me the account, one thing they've said or done recently with where you saw it and when, and your one-sentence value thesis." Build a minimal record, score it, mark fast path, continue.

## Step 1: Pick the path from readiness.md
Read the account's PAIN, MOMENT, CONFIDENCE and the message eligibility table in `os-decide/skills/os-account-scorer/references/readiness.md`.
- **Confirmed pain (PAIN 4 or 5):** state the problem in their words. At 5, name the consequence they named and make the case. At 4, ask about the consequence.
- **Signal only (PAIN 3, MOMENT ≥ 3, CONFIDENCE ≥ 3):** ask whether the suspected problem exists. Example: a new CRO and a CRM rollout support "Does your rollout include call-to-CRM handoffs?" They do not support "You're struggling with CRM hygiene." Never assert.
- **Insufficient (PAIN ≤ 2, or any supporting fact expired or synthetic):** write no message. Produce a research task: what to find, where, by when. Route to `os-account-context`.
Write the path and the reason at the top of the file.

## Step 2: The six beats
Signal (verified change, where seen, when) → Pain (their words, or the question on the signal-only path) → Consequence (theirs, never yours; omitted on signal-only) → Value (thesis applied) → Proof (one entry from the proof register) → Action (one small ask).

## Step 3: Persona map
Fill three roles from the record; find contacts in whatever CRM, email, or enrichment source is connected, or ask. Account facts stay constant across roles. The question and the next step change:
- Workflow owner: workload, process quality, adoption. Ask about the day-to-day.
- Economic buyer: business consequence, investment, success measures. Ask about the number.
- Technical owner: permissions, integration, maintenance. Ask about the stack.
Target three contacts by day 14.

## Step 4: Proof from the register only
Every proof point must exist in `context/messaging.md` under the proof register with: source, measurement period, scope, permission for external use. If the register lacks an entry, say "no approved proof for this claim" and leave the beat as a question. Never turn a workflow improvement into a revenue claim.

## Step 5: Formats
One email under 120 words, one LinkedIn note under 60, one call opener of three sentences, per role. Match tone rules.

## Step 6: Hand to the evaluator
Set `draft status: pending`. Do not mark approved. Say: "Run `os-message-evaluator` before anything leaves." If the user is running `os-act-loop`, it does this automatically.

## Quality gate
No message on the insufficient path. No consequence beat on the signal-only path. No proof outside the register.

What does the Os Message Architecture skill do?

Writes outreach for one account from its Account Context using the path its evidence allows (confirmed pain, signal-only question, or research task), mapped to three decision roles, with every proof point drawn from the proof register.

Where does the Os Message Architecture skill come from?

Built by Jonathan Kvarfordt. Standalone bundle v0.2.. The source and included files are shown on this page and in the package download.

Why was the Os Message Architecture skill chosen for this library?

It covers a job one of the eight revenue seats has to finish, and it states its process, decision rules, and output format clearly enough for an agent to follow.

When should the Os Message Architecture skill not be used?

Use it only for the job named above. A skill applied outside its scope produces confident output on the wrong question.

How do I install the Os Message Architecture SKILL.md file?

Download the file, create a folder named exactly os-message-architecture inside your agent's skills directory, and save the file inside it as SKILL.md. The agent loads it when a request matches the description.