Recommendation CanvasSAFE

Product Management skills framework built on battle-tested methods for Claude Code, Cowork, Codex, and AI agents.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: recommendation-canvas
argument-hint: "[AI product idea]"
description: Evaluate an AI product idea across outcomes, hypotheses, risks, and positioning. Use when deciding whether an AI solution deserves investment or recommendation.
intent: >-
  Evaluate and propose AI product solutions using a structured canvas that assesses business outcomes, customer outcomes, problem framing, solution hypotheses, positioning, risks, and value justification. Use this to build a comprehensive, defensible recommendation for stakeholders and decision-makers—especially when proposing AI-powered features or products that carry higher uncertainty and risk.
type: component
theme: validation-experiments
best_for:
  - "Deciding whether an AI product idea deserves real investment"
  - "Surfacing the risks and hypotheses behind an AI feature request"
  - "Comparing AI solution options on outcomes rather than novelty"
scenarios:
  - "Leadership wants an AI feature and I need to evaluate whether it's worth building"
  - "I have three AI solution options and need to compare them on outcomes and risk"
estimated_time: "30-45 min"
---


## Purpose
Evaluate and propose AI product solutions using a structured canvas that assesses business outcomes, customer outcomes, problem framing, solution hypotheses, positioning, risks, and value justification. Use this to build a comprehensive, defensible recommendation for stakeholders and decision-makers—especially when proposing AI-powered features or products that carry higher uncertainty and risk.

This is not a feature spec—it's a strategic proposal that articulates *why* this AI solution is worth building, *what* assumptions need validating, and *how* you'll measure success.

## Input

**Works best with:** The AI product or feature idea being evaluated.
**Also useful:** Target customer, expected business outcome, known risks, and who the recommendation must convince.

Anything supplied with the invocation itself — text after the skill name, a pasted context dump, or an appended `ARGUMENTS:` line — counts as answers already given. Use it and skip whatever it covers; don't re-ask.

**Arriving empty-handed? That works too.** The skill asks for the idea and the decision-maker, then works through the canvas boxes.

**Example invocation:** `Recommendation canvas: AI-suggested reorder quantities for warehouse managers — VP Ops wants a go/no-go next month.`

## Key Concepts

### The Recommendation Canvas Framework
Created for Dean Peters' Productside "AI Innovation for Product Managers" class, the canvas synthesizes multiple PM frameworks into one strategic view:

**Core Components:**
1. **Business Outcome:** What's in it for the business?
2. **Product Outcome:** What's in it for the customer?
3. **Problem Statement:** Persona-centric problem framing
4. **Solution Hypothesis:** If/then hypothesis with experiments
5. **Positioning Statement:** Value prop and differentiation
6. **Assumptions & Unknowns:** What could invalidate this?
7. **PESTEL Risks:** Political, Economic, Social, Technological, Environmental, Legal
8. **Value Justification:** Why this is worth doing
9. **Success Metrics:** SMART metrics to measure impact
10. **What's Next:** Strategic next steps

### Why This Works
- **Outcome-driven:** Forces clarity on business AND customer value
- **Hypothesis-centric:** Treats solution as a bet to validate, not a commitment
- **Risk-explicit:** Makes assumptions and risks visible upfront
- **Executive-friendly:** Comprehensive but structured for C-level review
- **AI-appropriate:** Especially useful for AI features with high uncertainty

### Anti-Patterns (What This Is NOT)
- **Not a PRD:** This is strategic framing, not detailed requirements
- **Not a business case (yet):** It informs the business case but needs validation first
- **Not a feature list:** Focus on outcomes, not capabilities

### When to Use This
- Proposing a new AI-powered product or feature
- Pitching to execs or securing budget/sponsorship
- Evaluating whether an AI solution is worth pursuing
- Aligning cross-functional stakeholders (product, engineering, data science, business)
- After completing initial discovery (you need context to fill this out)

### When NOT to Use This
- For trivial features (don't over-engineer small tweaks)
- Before any discovery work (you need user research and problem validation first)
- As a replacement for experimentation (canvas informs experiments, not vice versa)

---

## Application

Use `template.md` for the full fill-in structure.

### Step 1: Gather Context
Before filling out the canvas, ensure you have:
- **Problem understanding:** User research, pain points (reference `skills/problem-statement/SKILL.md`)
- **Persona clarity:** Who experiences the problem? (reference `skills/proto-persona/SKILL.md`)
- **Market context:** Competitive landscape, category positioning
- **Business constraints:** Budget, timelines, strategic priorities

**If missing context:** Run discovery work first. This canvas synthesizes insights—it doesn't create them.

---

### Step 2: Define Outcomes

#### Business Outcome
What's in it for the business? Use this format:
- [Direction] [Metric] [Outcome] [Context] [Acceptance Criteria]

```markdown
## Business Outcome
- [e.g., "Reduce by 25% the churn of existing customers using our existing product"]
```

**Example:**
- "Increase by 15% the monthly recurring revenue from enterprise customers within 12 months"

**Quality checks:**
- **Measurable:** Can you track this metric?
- **Time-bound:** Within what timeframe?
- **Ambitious but realistic:** Not "10x revenue in 1 month"

---

#### Product Outcome
What's in it for the customer? Use this format:
- [Direction] [Metric] [Outcome] [Context from persona's POV] [Acceptance Criteria]

```markdown
## Product Outcome
- [e.g., "Increase the speed of finding patients when I know the inclusion and exclusion criteria"]
```

**Example:**
- "Reduce by 60% the time spent manually processing invoices for small

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.