On Activation

1. Check if brand/ directory exists in the project root.
2. If it does, read available files: voice-profile.md, positioning.md, audience.md, creative-kit.md, stack.md, learnings.md.
3. Apply any loaded brand context to enhance output quality.
4. If brand/ does not exist, proceed without it — this skill works standalone.
5. Read brand/landscape.md if it exists — check the Claims Blacklist before making ecosystem or competitive claims. If stale (>14 days) or missing, warn that market claims may be outdated.

Note: Examples below use fictional brands (Acme, Lumi, Helm). Replace with your own brand context.

Positioning & Angles

The same product can sell 100x better with a different angle. Not a different product. Not better features. Just a different way of framing what it already does.

This skill finds those angles.

Brand memory integration

Reads: audience.md, competitors.md, voice-profile.md (if they exist)

On invocation, check for ./brand/ and load available context:

1. Check for ./brand/positioning.md -- if it exists, this is an update session:

   • Read the existing positioning file
   • Display the current primary angle and any saved alternatives
   • Ask: "You already have positioning on file. Do you want to refine it with fresh data, or start from scratch?"
   • "Refine" -- load existing angles, run competitive search for new data, suggest adjustments to current positioning based on what has changed in the market
   • "Start fresh" -- run the full process below as if no positioning exists

2. Load audience.md (if exists):

   • Use audience segments, pain points, and language patterns to inform angle generation
   • Show: "I see your audience profile -- [brief summary]. Using that to shape angles."

3. Load competitors.md (if exists):

   • Use known competitors as starting seeds for the competitive web search step
   • Show: "I found [N] competitors in your brand memory. Starting search from there."

4. Load voice-profile.md (if exists):

   • Use voice DNA to ensure angle language matches brand tone
   • Show: "Your voice is [tone summary]. Angles will match that register."

5. If ./brand/ does not exist:

   • Skip brand loading entirely. Do not error.
   • Note in opening message: "No brand profile found — this skill works standalone. I'll ask what I need. You can run /cmo or /brand-voice later to unlock personalization."

The core job

When someone asks about positioning or angles, the goal is not to find THE answer. It is to surface multiple powerful options they can choose from.

Every product has several valid angles. The question is which one resonates most with the specific audience at the specific moment.

Output format: 3-5 distinct angle options, numbered with circled numbers, each with:

• Statement (one sentence positioning)
• Psychology (why this works with this audience)
• Headline direction (how it would sound in copy)
• Best for (market conditions, audience segments)
• One option marked with ★ recommended

The angle-finding process

Step 1: Identify what they are actually selling

Not the product. The transformation.

Ask: What does the customer's life look like AFTER? What pain disappears? What capability appears? What status changes?

A fitness program does not sell workouts. It sells "fit into your old jeans" or "keep up with your kids" or "look good naked."

A SaaS tool does not sell features. It sells "close your laptop at 5pm" or "never lose a lead" or "stop the spreadsheet chaos."

The transformation is the raw material for angles.

Step 2: Map the competitive landscape

What would customers do if this did not exist? Not competitors -- alternatives.

• Do nothing (live with the problem)
• DIY (cobble together a solution)
• Hire someone (consultant, freelancer, agency)
• Buy a different category (different approach entirely)
• Buy a direct competitor

Each alternative has weaknesses. Those weaknesses become angle opportunities.

Angle opportunity: What is frustrating about each alternative that this solves?

Step 2.5: Competitive web search (live data)

Before generating angles, search the web for real competitor messaging. This grounds the angle work in current market reality rather than assumptions.

Search process:

1. Identify search targets:

   • If ./brand/competitors.md exists, start with those competitor names and URLs
   • If the user named competitors, search those
   • Otherwise, search for "[product category] + [target market]" to find the top players

2. Pull messaging data from competitor sites:

   • Homepage headlines and hero copy
   • Taglines and value propositions
   • Key claims on feature/pricing pages
   • Social proof framing (how they present testimonials)
   • CTA language

3. Map the landscape:

   • What claims appear on 3+ competitor sites (saturated territory)
   • What angles only 1 competitor uses (partially claimed)
   • What angles NO competitor uses (white space)
   • What proof/mechanism language dominates the space

4. Present findings as a competitive landscape map:

──────────────────────────────────────────────────

  COMPETITIVE MESSAGING LANDSCAPE

──────────────────────────────────────────────────

  Competitors Analyzed
  ├── [Competitor 1] -- "[their headline]"
  ├── [Competitor 2] -- "[their headline]"
  ├── [Competitor 3] -- "[their headline]"
  └── [Competitor 4] -- "[their headline]"

  ──────────────────────────────────────────────

  Saturated Claims (everyone says this)
  ├── "[Claim 1]"
  ├── "[Claim 2]"
  └── "[Claim 3]"

  Partially Claimed (1-2 competitors)
  ├── "[Claim]" -- used by [Competitor]
  └── "[Claim]" -- used by [Competitor]

  Underexploited Territory
  ├── Nobody is talking about [gap 1]
  ├── The [specific angle] is wide open
  └── [Niche audience] has no champion

──────────────────────────────────────────────────

Why this matters: Angles built on white space outperform angles that echo the market. If every competitor says "all-in-one platform," that phrase is dead. The competitive search reveals what NOT to say and where opportunity lives.

If web search is unavailable:

1. Use ./brand/competitors.md if it exists — extract messaging data from previous competitive intel.
2. If no competitors.md exists, ask the user to name 2-3 competitors and describe their positioning.
3. If no competitor data is available at all, skip the competitive landscape map and note the limitation: "Angles are not competitively grounded. Run /competitive-intel first for stronger differentiation."
4. Proceed to Step 3 with whatever context is available — the skill still works without competitive data, just at a lower enhancement level.

Step 3: Find the unique mechanism

The mechanism is HOW the product delivers results differently.

Not "we help you lose weight" (that is the promise). "We help you lose weight through intermittent fasting optimized for your metabolic type" (that is the mechanism).

The mechanism makes the promise believable. It answers: "Why will this work when other things have not?"

Questions to surface the mechanism:

• What is the proprietary process, method, or system?
• What do you do differently than the obvious approach?
• What is the counterintuitive insight that makes this work?
• What is the "secret" ingredient, step, or element?

Even if nothing is truly proprietary, there is always a mechanism. Name it.

Step 4: Assess market sophistication

Where is the market on Schwartz's awareness scale?

Stage 1 (New category): The market has not seen this before. Angle: Simple announcement. "Now you can [do thing]."

Stage 2 (Growing awareness): Competition exists, market is warming. Angle: Claim superiority. "The fastest/easiest/most complete way to [outcome]."

Stage 3 (Crowded): Many players, similar claims, skepticism rising. Ang