Feat/marketing os operating layer

ARCHITECTURE.md

@@ -50,6 +50,11 @@ All positioning skills load:
       - positioning/{source brand}
       - brand-voice/{source brand}
 
+    strategy/program-brief
+      - foundation/marketing-os
+      - marketing-science/measurement
+      - positioning/{relevant brand}
+
 ## Craft
 
 ### Core craft (not channel-specific)
@@ -189,3 +194,9 @@ Each tier orchestrates a specific set of craft skills:
     marketing-science/brand-equity
       - foundation/marketing-os
       (theoretical skill, anchored in Aaker/Keller/etc.)
+
+    marketing-science/measurement
+      - foundation/marketing-os
+      - marketing-science/brand-equity
+      (subscriber acquisition, cohorts, fame, authority, decision gates, and
+      compounding from results)

README.md

@@ -2,7 +2,8 @@
 
 A skills repository for Every's marketing OS. Each skill is a structured prompt module designed to be loaded into Plus One (or any Claude-based workflow) for a specific marketing task.
 
-This repo is **scaffolding only** — stubs will be populated as the corresponding brand docs and frameworks are completed.
+The repo combines active operating skills with scaffolding that is populated as Every's
+brand documents, measurement systems, and channel strategies are finalized.
 
 ## Taxonomy
 
@@ -11,10 +12,10 @@ The repository is organized into seven top-level folders:
 1. **foundation/** — The Marketing OS, the operating system underlying all skills
 2. **brand-voice/** — Voice guides for Every's master brand and each sub-brand (Cora, Spiral, Monologue, Plus One, Proof, Sparkle)
 3. **positioning/** — Positioning frameworks for Every master and each sub-brand
-4. **strategy/** — Higher-order strategy skills (messaging architecture, one-pagers) that compose voice and positioning
+4. **strategy/** — Higher-order strategy skills (messaging architecture, program briefs, one-pagers) that compose voice, positioning, and measurement
 5. **craft/** — Execution-level skills for specific deliverables: copywriting, editing, naming, and channel-specific output (launch emails, LinkedIn posts, X posts, website copy, press comms)
 6. **launches/** — Orchestration skills for three launch tiers (improvement, feature, new product) that load strategy and all channel craft skills
-7. **marketing-science/** — Research, archetyping, and brand-equity theory skills that inform positioning and naming
+7. **marketing-science/** — Research, archetyping, brand equity, and measurement skills that inform positioning, programs, and decisions
 
 ## Dependency model
 
@@ -24,4 +25,6 @@ See [ARCHITECTURE.md](ARCHITECTURE.md) for the full dependency graph.
 
 ## Status
 
-All 31 skill stubs are in place. Each will be populated as its corresponding brand doc or framework is finalized.
+The foundation doctrine, Every-master positioning, audience model, measurement system,
+program briefs, and core editing workflows are active. Some product-level voice,
+positioning, and channel skills remain scaffolding.

craft/editing/SKILL.md

@@ -33,13 +33,18 @@ Does this copy actually work commercially? The final filter — evaluating wheth
 copy does what it was written to do (convert, persuade, inform, engage) rather than
 just sounding good.
 
+### references/compound-from-edits.md
+How to learn from Douglas's edits: compare versions, classify the delta, preserve his
+version as canonical, and update the relevant durable rule without overwriting the work.
+
 ## When to invoke
 
 - When reviewing and refining a draft for voice alignment
 - When tightening copy that's already been written
 - When a piece needs a final pass before publication
 - When evaluating whether existing copy meets brand and effectiveness standards
 - When stripping AI tells from generated or co-written copy
+- When Douglas returns an edited draft and asks to compound or remember the changes
 
 ## Reference routing
 
@@ -49,13 +54,17 @@ just sounding good.
 | Detecting or removing AI-generated patterns | `references/ai-tells.md` |
 | Structural editing decisions (cut, rewrite, reorder) | `references/surgery.md` |
 | Evaluating commercial effectiveness of copy | `references/effectiveness-check.md` |
+| Learning durable rules from a human-edited draft | `references/compound-from-edits.md` |
 
 ## Dependencies
 
 - `foundation/marketing-os`
 - `brand-voice/{relevant brand}` (loaded dynamically)
 
-## Status
+## Quick checklist
 
-Reference skeletons are in place. Filters and AI-tells will be populated first; surgery
-and effectiveness-check will follow.
+- Did the edit preserve the argument and creative premise?
+- Did it remove repetition rather than unique information?
+- Is the language active, specific, and recognizably Every?
+- Does the piece still do its intended job?
+- If Douglas edited it, did the system learn from the delta?

craft/editing/references/compound-from-edits.md

@@ -0,0 +1,38 @@
+# Compound From Edits
+
+Human edits are evidence about the system, not cleanup to be absorbed silently.
+
+## Workflow
+
+1. Preserve the AI draft and the edited version.
+2. Compare them before producing another rewrite.
+3. Classify each meaningful change:
+   - Strategic correction.
+   - Stronger point of view.
+   - Added specificity or cultural texture.
+   - Active voice or sharper prose.
+   - Better structure or hierarchy.
+   - Stronger evidence.
+   - Clearer ownership or measurement.
+   - Formatting or native-document preservation.
+4. Identify repeated patterns across the edit.
+5. Update the narrowest durable source: brand voice, positioning, audience, measurement,
+   channel guidance, collaboration rule, or project context.
+6. Treat the human-edited version as canonical.
+
+## Guardrails
+
+- Do not rewrite the whole artifact while extracting the lesson.
+- Do not convert a project-specific preference into a universal rule.
+- Do not infer that every deletion is a general ban; determine the reason.
+- Preserve disagreement and unresolved choices as project context, not doctrine.
+- State what changed and why in plain language.
+
+## Output
+
+Return:
+
+- The three to seven most important changes.
+- Why each improved or redirected the work.
+- The durable rule or file to change.
+- Any project-specific context to keep local.

craft/editing/references/surgery.md

@@ -1,24 +1,48 @@
-# Surgery — Cut vs. Rewrite vs. Reorder
+# Surgery: Cut, Rewrite, or Reorder
 
-The decision framework for structural editing — knowing whether a paragraph needs a
-scalpel, a full rewrite, or a different position in the piece.
+## Cut
 
----
+Cut repetition, throat-clearing, empty theory, stale context, generic claims, and process
+detail that does not affect the reader's decision.
 
-## When to cut
-<!-- To populate — signals that content should be removed entirely -->
+Do not cut a unique audience insight, mechanism, proof point, owner, decision, metric,
+cultural reference, or creative detail merely to reduce word count.
 
-## When to rewrite
-<!-- To populate — signals that the idea is right but the execution is wrong -->
+## Rewrite
 
-## When to reorder
-<!-- To populate — signals that the structure is wrong but the content is sound -->
+Rewrite when the idea is right but the sentence is passive, abstract, generic, overly
+theoretical, or mechanically awkward. Preserve the claim and strengthen its form.
 
-## When to merge
-<!-- To populate — signals that two sections are saying the same thing differently -->
+Prefer:
 
-## When to leave it alone
-<!-- To populate — signals that intervention would make things worse -->
+- "Every does" to "Every should do."
+- Named actors and actions to passive constructions.
+- Concrete mechanisms to strategic abstractions.
+- One sharp example to several loose ones.
 
-## Decision framework
-<!-- To populate — a structured approach to choosing the right surgical intervention -->
+## Reorder
+
+Reorder when the decision, argument, or recommendation appears after its supporting
+context. Lead with the point, then earn it.
+
+## Merge
+
+Merge sections only when they perform the same job. Keep them separate when they address
+different audiences, mechanisms, phases, owners, or decisions.
+
+## Leave it alone
+
+Leave language alone when an unusual rhythm, cultural reference, material detail, or
+provocation gives the idea its identity. Editorial smoothness is not always an
+improvement.
+
+## Structural test
+
+Before finalizing, ask:
+
+1. What unique information existed before the edit?
+2. Where does each item live now?
+3. What became clearer?
+4. What became more generic?
+
+Reverse any edit whose main effect was making the work more generic.

craft/press-comms/SKILL.md

@@ -1,12 +1,36 @@
 # Press Comms
 
-Generates PR pitch emails based on product details, target outlet/journalist, and angle — crafted to land coverage.
+Builds Every's press strategy and writes the materials required to earn coverage. Press
+combines an always-on expert-commentary desk with concentrated tentpole launches.
 
 ## When to invoke
 
 - When pitching a journalist or outlet on a product launch or story
 - When a launch skill orchestrates channel execution and needs the press deliverable
 - When drafting a standalone PR pitch for a specific media target
+- When building a press plan, spokesperson strategy, exclusive, press kit, or quarterly
+  PR readout
+
+## Operating model
+
+### Always-on press
+
+Maintain a steady cadence of timely expert commentary, relationship building, and
+briefings. The operating target is at least one meaningful expert feature each month.
+Dan is the lead public voice; the wider roster expands the number of credible stories
+Every can enter.
+
+### Tentpole press
+
+Reserve the full launch machine for news that can change how the market sees Every.
+Choose one major outlet and reporter whose audience and beat fit the story. Offer a
+time-bound exclusive with early access, a private demonstration, the relevant
+spokesperson, and the strongest evidence Every can share. Once the exclusive publishes,
+release Every's definitive launch story and open distinct angles to the broader press
+list.
+
+Load `references/every-pr-strategy.md` for the current spokesperson, outlet, partnership,
+and measurement model.
 
 ## Dependencies
 
@@ -21,9 +45,19 @@ Brand is resolved dynamically based on what invokes this skill.
 ## Channel-specific notes
 
 - Every pitch is tailored to a specific journalist and outlet — no blast templates
-- The angle must match the outlet's editorial lens, not just the company's messaging
+- The angle must match the outlet's editorial lens rather than repeat the company's
+  messaging
 - Subject line and opening sentence determine whether the pitch gets read
-
-## To do
-
-Populate with press comms guidance once PR strategy work is complete.
+- Impressions describe potential exposure; they do not prove impact
+- Every does not pay media companies for coverage
+- Paid partnerships are acceptable when the partner pays Every, the commercial
+  relationship is clear, and Every retains editorial and creative control
+
+## Quick checklist
+
+- Is the story genuinely newsworthy for this reporter's audience?
+- Does the pitch lead with the story rather than company background?
+- Is the spokesperson matched to the subject?
+- Are claims, methodology, screenshots, and proof ready?
+- Is the exclusive narrow enough to preserve wider launch angles?
+- Does the plan include authority, attention, acquisition, and business measurement?

craft/press-comms/references/every-pr-strategy.md

@@ -0,0 +1,62 @@
+# Every PR Strategy
+
+## Objective
+
+Make Every a recurring expert source on AI and work, led by Dan Shipper, while using
+tentpole launches, research, partnerships, and company practitioners to broaden the
+company's authority.
+
+## Spokespeople
+
+- **Dan Shipper:** priority voice for broad audiences; leading expert on AI and work,
+  known for philosophical takes, accessible explainers, and deep working knowledge.
+- **Brandon Gell and Kate Lee:** company, Editorial, and business stories where relevant.
+- **Douglas Brundage:** marketing, brand, culture, and creative strategy.
+- **Austin and Yash:** Growth and go-to-market.
+- **Natalia Karbasova and Mike Knoop:** AI transformation and Consulting.
+- **Kieran Klaassen and Willie Stillwell:** engineering and technical practice.
+
+The spokesperson roster supports Dan. It does not dilute his lead role.
+
+## Media priorities
+
+- Business and mainstream: Bloomberg, The Wall Street Journal, Financial Times, The New
+  York Times, CNBC, and Fortune.
+- Technology and AI: Wired, The Verge, Axios, Semafor, Fast Company, Business Insider,
+  and TechCrunch.
+- High-trust shows and newsletters where long-form expertise matters.
+- Specialist engineering, Consulting, design, marketing, and media outlets matched to
+  the spokesperson.
+
+Maintain a named reporter list by beat, relationship, and story fit. Pitch individually.
+
+## Operating cadence
+
+- Run a short weekly press desk to select timely stories and owners.
+- Maintain same-day expert availability for rapid response when schedules permit.
+- Keep a source bank of company facts, approved claims, research, customer proof,
+  biographies, and points of view.
+- Brief priority reporters between launches.
+- Use smaller updates to feed commentary, owned Editorial, podcasts, and social. Do not
+  force each update into a standalone announcement.
+
+## Partnerships
+
+Pursue sponsor-funded reports, shows, events, or physical experiences when the format
+strengthens Every's authority. Every retains its point of view, labels the commercial
+relationship, and protects Editorial independence.
+
+## Measurement
+
+Report quarterly:
+
+- **Authority:** outlet quality, prominence, message pull-through, repeat relationships,
+  expert invitations, and citations.
+- **Attention:** estimated impressions, branded search, direct traffic, social sharing,
+  and Fame Score movement.
+- **Acquisition:** press-attributed free and paid subscribers, conversion, retention,
+  and fully loaded CAC when tracking permits.
+- **Business:** Consulting inquiries, qualified opportunities, influenced pipeline, and
+  partner interest.
+
+Never convert coverage into fictional advertising value.

craft/social/instagram-post/SKILL.md

@@ -1,12 +1,27 @@
 # Social — Instagram Post
 
-Writes captions and copy for Instagram posts, carousels, and Reels.
+Builds and executes Every's Instagram editorial strategy across Reels, carousels,
+Stories, captions, collaborations, and launch moments. The account turns visual discovery
+into free subscriptions without becoming a promotional feed.
 
 ## When to invoke
 
 - When drafting Instagram captions for a launch, brand moment, or content series
 - When a launch skill orchestrates channel execution and needs the Instagram deliverable
 - When writing copy for carousels, Reels overlays, or Stories
+- When planning the Instagram channel, content franchises, publishing cadence,
+  collaborator strategy, or acquisition pilot
+
+## Every channel promise
+
+**Follow Every to stay at the edge of AI. Subscribe to become AI native.**
+
+Every runs an editorial account. One recurring on-camera anchor leads the first six
+months. Dan launches the account and appears at major moments; employees and outside
+practitioners join as beat-specific experts.
+
+Load `references/every-instagram-strategy.md` for content franchises, format roles,
+collaborations, handles, and measurement.
 
 ## Dependencies
 
@@ -21,9 +36,17 @@ Brand is resolved dynamically based on what invokes this skill.
 ## Channel-specific notes
 
 - Visual-first platform — copy supports the image, not the other way around
-- 2,200-character caption limit but engagement drops after ~125 characters above fold
-- Hashtag strategy and alt text are part of the deliverable
-
-## To do
-
-Populate with Instagram post guidance once social strategy work is complete.
+- Reels drive discovery, carousels earn saves, Stories build habit, and collaborator
+  posts bring established audiences
+- Every editorializes AI in real time and translates its own work into visual-native
+  formats
+- The subscriber path and production cost are part of the deliverable
+
+## Quick checklist
+
+- Does the post make a Seeker feel closer to the edge of AI?
+- Is there editorial judgment rather than promotional repackaging?
+- Is the format native to Reels, carousels, or Stories?
+- Does the visual idea carry the post before the caption?
+- Is the collaborator genuinely relevant?
+- Is the free-subscriber path tracked?