← ALL SKILLS

message-house

TRIGGERS ON

When the user wants to build a message house, convert positioning into a narrative architecture, create messaging pillars, draft proof points, or generate channel-mapped messaging lines. Also triggers on 'we need a message house', 'turn this positioning into messaging', 'build the pillars', 'what's our narrative', or 'we keep saying different things across channels'.

INSTALL
/plugin marketplace add Scylark/manual-focus
/plugin install the-lens@manual-focus

Message house

You convert a positioning brief into a four-tier message house: narrative at the top, 3–4 pillars beneath, proof points beneath those, channel-specific lines at the floor. Each pillar is claim-shaped, not organisational. Each proof point traces to a source.

Inputs to gather first

  1. Positioning brief — read from .lens/positioning-brief.md if it exists, or ask for the equivalent. Must include the one-sentence position, 3–5 proof points, audience definition, what-not-this. Refuse to start if the positioning is vague.
  2. Channels to map messaging against — paid ads, social, email, landing pages, sales pitch
  3. Voice profile — read from .lens/voice-profile.json if it exists, or ask. Channel lines need to pass the voice rubric.

The pipeline (four phases)

Phase 1 — Inputs validation

Confirm the positioning brief is sharp. Test:

  • Can you state the brand’s position in one sentence?
  • Are the proof points concrete (data, customer story, partnership, technical claim)?
  • Is the audience defined as a behavioural segment, not just a demographic?
  • Does the brief name what the brand is explicitly NOT?

If any test fails, send the brief back. Vague positioning produces a vague message house.

Phase 2 — Narrative drafting

Generate the one-sentence brand narrative that subsumes the positioning. Then propose 2–3 alternative pillar structures, each with 3–4 candidate pillars. Pillars must be claim-shaped:

  • Bad pillar: “Product” (category, not claim)
  • Good pillar: “The work compounds because we own the full pipeline” (claim, falsifiable, defensible)

Present the alternatives. User picks one structure.

Phase 3 — Proof point expansion

For each pillar, surface 3–5 proof points. Each gets:

  • One-sentence summary
  • Source tag (data / customer / partnership / technical / market)
  • Reusability flag (one-time use vs evergreen)
  • Verification status: verified (checked against a primary source), source_supplied (taken from the brand’s own site or deck, not independently checked), pending, or invented_DO_NOT_USE. Never invent a statistic to fill a gap.

Customer story proof points require active customer permission to publish. Flag any that need re-confirmation.

Phase 4 — Channel-mapped lines

For each pillar, generate channel-specific lines:

ChannelConstraintForm
Landing-page H112 words maxOutcome-anchored
Email subject60 chars maxSpecific not generic
Social post (LinkedIn)200 wordsHook on line 1
Social post (X)280 charsArgument-compressed
Paid ad headline30 charsBenefit-led
Sales talking point3 bulletsPillar + proof + ask

Punctuation follows the voice profile, not a house default. Sales talking points keep the qualifiers (exclusions, caps, time windows) the source proof carries.

All lines run through the lexical checks of the voice rubric (banned words, punctuation, contractions). Lines that fail go back for regeneration.

Output

Single document with these sections:

  1. The narrative — one sentence, the top of the house
  2. The pillars — 3–4 named, each with one-sentence definition
  3. The proof — proof points per pillar, with sources
  4. The lines — channel-mapped messaging per pillar
  5. The rebuttal sheet — 8–12 pushback questions buyers / press / internal stakeholders are likely to ask, with answers in pillar language. Cover price, fit and category at minimum. If no verified proof answers a pushback, mark it PROOF GAP rather than stretching another proof to fit

Save to .lens/message-house.md. Downstream skills (lifecycle-journey-builder, social-content-factory, earned-media-pitch, eval-gated-drafting) read this for messaging consistency.

Evals

Self-check before delivery:

  • Pillar distinctiveness — pairwise semantic similarity between pillars <0.7. If two pillars are saying the same thing, merge or sharpen.
  • Proof traceability — every proof point has a source tag. Unsourced claims don’t ship.
  • Channel-line voice match — every line passes the lexical checks of the voice rubric (banned words, punctuation, contractions). Sentence-length, variance and paragraph checks don’t apply to lines under 280 characters. If a line passes on the website but fails in social, flag — the brand may have two voices.

Failure modes to watch

  • Pillars become categories — “Product / People / Process” is organisational, not claim-shaped. Pillars must be arguments the brand is making.
  • Proof points get reused without permission — customer quotes require active permission. The brand-friendly version is not the authorised version.
  • Channel lines drift over time — message house is a living document. Quarterly refresh of channel lines; narrative and pillars stay stable.
  • Trying to be too many things — a four-pillar house is the ceiling. Five+ pillars means the positioning isn’t focused. Push back to the positioning brief.

Hand-off

When done, the message house is the source of truth for all subsequent content. Downstream skills will reference it:

  • lifecycle-journey-builder uses the channel lines for touchpoint copy
  • earned-media-pitch uses the narrative and pillars for journalist framing
  • eval-gated-drafting uses the pillars to score whether a piece advances the position
THE LENS

Twenty-eight skills, forty-seven playbooks, growing.

Browse the rest of the skill set or read the paired playbook for the strategic context.