brand-identity-refinement

What this skill does

# brand-identity-refinement

Refine brand identity through structured metaphor testing, competitive research, sentiment analysis, and owner-grounded verification — producing a two-layer identity document where every claim is traceable to current repository reality.

## Triggers

Alternate expressions and non-obvious activations (primary phrases are matched automatically from the skill description):

- "brand refresh" → full identity refinement cycle
- "what's our story" → identity seed intake + metaphor generation
- "refine our positioning" → metaphor testing and two-layer document generation
- "brand session" → guided owner contribution + structured refinement
- "fix the brand" → claim audit + identity realignment
- "brand voice doesn't feel right" → structural metaphor re-evaluation

## Purpose

Brand identity work fails in predictable ways: it starts with what sounds good rather than what is true, commits to metaphors competitors already own, and produces claims the product cannot actually support. This skill exists to prevent all three failure modes.

It grounds identity work in repository reality before any creative output is produced. It tests metaphors against structural criteria rather than vibes. It researches competitive ownership of candidate metaphors. It creates explicit moments for domain expertise to correct AI assumptions. And it produces a two-layer document — metaphor front door, literal product room — so identity is both memorable and defensible.

The result is brand identity that a founder can say out loud and mean, that a salesperson can use without qualification, and that a skeptical prospect cannot immediately disprove.

## Behavior

When triggered, this skill executes in eight ordered steps:

### Step 1: Ground in Reality

Before any creative work begins, read the codebase and documentation to establish a verified capability baseline.

1. Identify all relevant repositories from the project context or user input
2. Read every top-level README and any linked capability documentation
3. Extract every explicit capability claim: what the system does, what it produces, what it integrates with, what it does not do
4. Build a "what is actually true right now" baseline document
5. If an existing identity document or brand brief is present, cross-check its claims against the extracted baseline
6. Flag any claim in the existing identity that the repos do not support — mark as `[UNVERIFIED]` with a note on what would be needed to substantiate it
7. Flag any capability in the repos that the existing identity does not mention — mark as `[UNCLAIMED ASSET]`

Output: `.aiwg/marketing/brand/capability-baseline.md`

### Step 2: Input Identity Seed

Accept the starting material for the refinement session.

1. Request or locate one of: existing identity document, brand brief, seed positioning statement, or founder description of the product
2. From the input, extract:
   - Current structural metaphor (if any — what frame does the language imply?)
   - Key claims being made
   - Target audience as currently described
   - Emotional register (aspirational, pragmatic, urgent, playful, authoritative?)
   - What the identity document implies the product *is for*
3. If no seed material is available, prompt the owner with: "Describe what this product does in one sentence, as if explaining to a smart person who has never heard of it."
4. Document the extracted seed elements alongside the capability baseline

Output: append to `.aiwg/marketing/brand/capability-baseline.md` under "Identity Seed" section

### Step 3: Structural Metaphor Testing

For the current metaphor and any proposed alternatives, apply a four-criteria evaluation before committing to any metaphor direction.

**Criterion 1 — Output capture**: Does the metaphor describe what the system *produces*, not just how it *operates*? A metaphor that describes mechanism but not outcome will feel abstract to buyers.

**Criterion 2 — Human role accuracy**: Does the human role implied by the metaphor match what users actually experience? If the metaphor casts the user as a passive recipient but the product requires active configuration, the metaphor will feel dishonest over time.

**Criterion 3 — Memory/persistence mapping**: Does the metaphor accommodate the product's memory, state, or continuity story naturally? Or does it require a forced extension that feels bolted on?

**Criterion 4 — Emotional register fit**: What is the emotional register the metaphor evokes, and does it match both the product's actual stakes and the target audience's self-image? A metaphor with the wrong emotional register will attract the wrong buyers and alienate the right ones.

For each candidate metaphor, score each criterion: Strong / Acceptable / Weak / Disqualifying. A single Disqualifying score eliminates the metaphor. Two Weak scores require a substantial reason to proceed.

Document results in a comparison table. Carry forward only Strong or Acceptable metaphors.

Output: `.aiwg/marketing/brand/metaphor-testing.md`

### Step 4: Competitive and Sentiment Research

Metaphors that pass structural testing must still clear two external checks.

**Competitive ownership check**: Research whether candidate metaphors are already strongly associated with a competitor, a category leader, or a broader cultural frame that would make the metaphor confusing or derivative. A metaphor you did not invent but that is not yet owned is viable. A metaphor already associated with a specific brand in the same category is not.

For each surviving metaphor:
- Identify the top 5–8 competitors and category-adjacent brands
- Search for metaphor usage in their positioning, taglines, and headline copy
- Rate metaphor availability: Unclaimed / Lightly used / Contested / Owned by competitor

**Sentiment and cultural context check**: Logical correctness does not guarantee cultural safety. Research current sentiment and connotation around each metaphor:
- Is there a recent cultural event or controversy that has changed the resonance of this metaphor?
- Does the metaphor carry unintended connotations for any segment of the target audience?
- Is the metaphor on an upward or downward cultural trajectory?

Eliminate any metaphor rated "Owned by competitor" or carrying a significant negative connotation. Flag contested metaphors for owner decision.

Output: `.aiwg/marketing/brand/competitive-landscape.md` (metaphor section appended to any existing competitive landscape file)

### Step 5: Domain Expertise Correction

This step creates a structured pause for the product owner's domain knowledge to correct AI-generated assumptions. It is not optional and cannot be synthesized from other sources.

Present findings so far to the owner and ask three explicit questions:

**Question 1 — Assumption audit**: "What assumptions am I making about the audience and their stakes that your domain expertise would correct? What do I not understand about how this product gets used in practice?"

**Question 2 — Internal phenomenology**: "What does it feel like to use this system when it's working well? Not what it produces — what it feels like from the inside."

**Question 3 — Owner completion prompt**: "Complete this sentence from your perspective, without editing for polish: '[Product] is the ___ for ___.' Say what first comes to mind."

Record the owner's responses verbatim. These responses take precedence over metaphors that passed all structural and competitive criteria. A metaphor the owner cannot say out loud is not a viable brand metaphor regardless of its score.

If the owner's completion prompt surfaces a metaphor not previously considered, add it to the candidate set and run it through Steps 3 and 4 before proceeding.

Output: owner responses recorded in `.aiwg/marketing/brand/owner-contributions.md`

### Step 6: Generate Two-Layer Identity

For each major identity element — positioning statement, tagline, elevator pitch, about description, mission statement — pro