Pi Strategy

A skill pack for the Pi Coding Agent that turns raw ideas into a practical, testable product strategy before build work.

It helps technical builders avoid overbuilding by forcing clarity on market, user, problem, offer, MVP, proof, messaging, and acquisition.

MARKET SIGNAL → USER CONTEXT → PROBLEM/OUTCOME → OFFER → MECHANISM → MVP → PROOF → MESSAGE → ACQUISITION → STRATEGY BRIEF

Why use this

AI can accelerate execution, but it can also accelerate waste.

Pi Strategy enforces a simple discipline:

1. define who you serve;
2. define the painful problem and desired outcome;
3. define a credible offer;
4. define the smallest MVP that proves the offer;
5. define proof and acquisition path before downstream execution.

Core rule:

Do not build before strategy.

Who this is for

• Technical founders
• Indie builders
• Small early-stage product teams
• Anyone repeatedly shipping without clear validation

Installation

From npm:

pi install npm:@felipefontoura/pi-strategy

Then reload Pi:

/reload

Available skills

Use through Pi skill commands:

/skill:strat-init
/skill:strat-market-signal
/skill:strat-user-context
/skill:strat-problem-outcome
/skill:strat-product-offer
/skill:strat-unique-mechanism
/skill:strat-mvp-scope
/skill:strat-proof-system
/skill:strat-message-system
/skill:strat-acquisition-path
/skill:strat-brief

/skill:strat-init

Initializes strategy workspace in the user project.

Typical output:

.ai/strategy/
.ai/strategy/index.md

/skill:strat-market-signal

Checks if there is enough reachable market signal to proceed.

Typical output:

.ai/strategy/signal/market-signal.md

/skill:strat-user-context

Defines concrete user and buyer context, workflow, triggers, and objections.

Typical output:

.ai/strategy/signal/user-context.md

/skill:strat-problem-outcome

Separates surface vs deeper problem and defines measurable desired outcome.

Typical output:

.ai/strategy/signal/problem-outcome.md

/skill:strat-product-offer

Builds the offer using the value equation.

Typical output:

.ai/strategy/offer/product-offer.md

/skill:strat-unique-mechanism

Defines why this works differently from the old way.

Typical output:

.ai/strategy/offer/unique-mechanism.md

/skill:strat-mvp-scope

Defines the smallest deliverable that proves the offer and mechanism.

Typical output:

.ai/strategy/offer/mvp-scope.md

/skill:strat-proof-system

Maps available proof, missing proof, and proof acquisition plan.

Typical output:

.ai/strategy/offer/proof-system.md

/skill:strat-message-system

Turns strategy into clear technical messaging.

Typical output:

.ai/strategy/gtm/message-system.md

/skill:strat-acquisition-path

Chooses the right CTA and conversion path based on readiness and trust.

Typical output:

.ai/strategy/gtm/acquisition-path.md

/skill:strat-brief

Consolidates all artifacts into final handoff.

Typical output:

.ai/strategy/handoff/strategy-brief.md

Recommended quickstart

/skill:strat-init
/skill:strat-market-signal
/skill:strat-user-context
/skill:strat-problem-outcome
/skill:strat-product-offer
/skill:strat-unique-mechanism
/skill:strat-mvp-scope
/skill:strat-proof-system
/skill:strat-message-system
/skill:strat-acquisition-path
/skill:strat-brief

Artifact structure

Strategy state lives in the user project:

.ai/strategy/
  index.md
  signal/
    market-signal.md
    user-context.md
    problem-outcome.md
  offer/
    product-offer.md
    unique-mechanism.md
    mvp-scope.md
    proof-system.md
  gtm/
    message-system.md
    acquisition-path.md
  handoff/
    strategy-brief.md

Optional durable context:

.ai/steering/

Handoff contract

Pi Strategy writes:

.ai/strategy/handoff/strategy-brief.md

Downstream packages should consume this file instead of inventing strategy again.

Ready for Handoff should only be used when:

• offer is specific and testable;
• MVP scope is bounded;
• proof exists or proof plan exists;
• message system is specific;
• acquisition path is coherent and measurable.

What this package does not do

It does not create implementation code, UI, ad systems, CRM automations, or full downstream execution assets.

Language policy

Package source/docs are EN-US.

At runtime, skills adapt to user language, while preserving canonical technical identifiers (file paths, status values, commands, and stable headings).

License

MIT