Value Proposition Generator
Analyzes your codebase to generate value propositions, elevator pitches, landing page hero copy, and before/after comparisons grounded in what the product actually does.
SKILL.md
---
description: Generate value propositions grounded in what the product actually does
allowed-tools: Read, Glob, Grep
---
# Value Proposition Generator
Analyze the codebase to understand what the product does, then generate multiple value proposition frameworks grounded in real functionality.
The product or feature to focus on: $ARGUMENTS
## Steps
1. **Discover product context** by reading key files:
- Read `README.md` or `README` for the product overview.
- Read `package.json`, `Cargo.toml`, `pyproject.toml`, or equivalent for metadata (name, description, keywords).
- Glob for `**/landing*`, `**/marketing*`, `**/about*` to find any existing positioning.
- Grep for common value-related terms: "benefit", "save", "fast", "easy", "secure", "automate".
2. **Analyze core functionality** to understand what the product actually does:
- Glob for route handlers, API endpoints, or CLI commands to catalog capabilities.
- Read the main entry point(s) to understand the primary workflow.
- Grep for integration points (third-party APIs, databases, services) to understand the ecosystem.
3. **Identify the target user** by examining:
- Authentication/authorization patterns (B2B vs B2C signals).
- UI copy, form labels, or CLI help text for persona clues.
- Configuration options that reveal use cases.
4. **Generate the Value Proposition Canvas**:
- **Customer Jobs**: What tasks is the user trying to accomplish?
- **Pains**: What frustrations exist with current solutions?
- **Gains**: What positive outcomes does the user want?
- **Pain Relievers**: How does this product address each pain?
- **Gain Creators**: How does this product deliver each gain?
- **Products & Services**: What specific features map to each?
5. **Generate positioning frameworks**:
**One-Liner** (under 10 words):
`[Product] helps [audience] [achieve outcome]`
**Elevator Pitch** (30 seconds / ~75 words):
Structure: Problem → Solution → Key Differentiator → Call to Action
**Landing Page Hero Copy**:
- Headline (benefit-driven, under 10 words)
- Subheadline (how it works, one sentence)
- Three bullet points (features framed as benefits)
- CTA button text
**Before/After Comparison**:
| Without [Product] | With [Product] |
Show 4-5 rows contrasting the painful status quo with the improved state.
6. **Generate messaging variations** for different channels:
- Twitter/X post (under 280 chars)
- LinkedIn post (professional tone, 2-3 paragraphs)
- Product Hunt tagline (under 60 chars)
## Rules
- Every claim must be traceable to actual product functionality found in the code.
- Never invent features that do not exist in the codebase.
- Use specific, concrete language — not generic marketing fluff.
- If $ARGUMENTS specifies a particular feature, focus the value propositions on that feature rather than the whole product.
- If the codebase lacks enough context, say so and note which files would help.How It Works
Most marketing copy is written by people who roughly understand what a product does. This skill flips that — it reads the actual source code and derives positioning from real functionality. The result is value propositions that are specific rather than generic, because they are grounded in what the code actually implements.
The Value Proposition Canvas output is particularly useful for founders and product managers. By mapping real features to customer jobs, pains, and gains, it surfaces positioning angles that might not be obvious from the outside. A feature you built for technical reasons might solve a pain point you never explicitly marketed.
The multi-format output (one-liner, elevator pitch, hero copy, social posts) means a single run produces copy for every channel. The before/after comparison is especially effective for landing pages because it lets visitors immediately see themselves in the "before" column and desire the "after" state.