# Canvas

> >

- **Type:** Skill
- **Install:** `agentstack add skill-panaversity-agentfactory-business-plugins-canvas`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [panaversity](https://agentstack.voostack.com/s/panaversity)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** Apache-2.0
- **Upstream author:** [panaversity](https://github.com/panaversity)
- **Source:** https://github.com/panaversity/agentfactory-business-plugins/tree/main/innovation/skills/canvas
- **Website:** https://agentfactory.panaversity.org/docs/Business-Domain-Agent-Workflows

## Install

```sh
agentstack add skill-panaversity-agentfactory-business-plugins-canvas
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

## CONTEXT LOADING

Before executing, check for `innov.local.md` in the working directory.
If found, extract:

- venture: name, stage, type, problem_statement, target_customer
- business_model_canvas: current version with all blocks and evidence quality
- customer_profiles: personas, pains, gains, WTP
- key_assumptions: business model assumptions

If `innov.local.md` is not found:
Continue with conversation context. After first substantive output, prompt:
"I'm working without your venture context. Run Exercise 8 from Chapter 40
to build innov.local.md -- it will make every subsequent output specific
to your venture rather than generic."

## STAGE-AWARE CALIBRATION

Check venture.stage and calibrate:

- IDEA: Warning -- "Building a canvas at the IDEA stage is premature. Most blocks will be ASSUMED with no evidence. Consider running /discovery first to get real customer data before designing a business model."
- DISCOVERY: A canvas is useful here -- you can start mapping hypotheses as you learn from customers.
- VALIDATION: This is your focus stage. The canvas should reflect what you are testing and learning.
- MVP: This is your focus stage. The canvas should be updated with pilot data and evidence.
- GROWTH: The canvas should be largely validated. Focus on optimising blocks, not hypothesising.

## DLA PROGRESSION CHECK

If no customer_profiles or discovery data exist in innov.local.md:
"You are building a business model canvas without customer discovery data.
Every block will be hypothetical. Consider running /discovery first --
a canvas grounded in customer evidence is vastly more useful than one
built from assumptions alone."

## BUSINESS MODEL CANVAS WORKFLOW

### Task Types

TYPE 1: CANVAS BUILD (first draft)
Input: Venture context from innov.local.md + any additional context
Output: All 9 blocks with hypotheses, evidence quality, biggest risk, open question

TYPE 2: CANVAS STRESS TEST
Input: Existing canvas
Output: One adversarial question per block + strongest counter-argument

TYPE 3: ALTERNATIVE MODELS
Input: Current canvas
Output: 3 alternative business model configurations with trade-off analysis

TYPE 4: CANVAS UPDATE (post-learning)
Input: Previous canvas version + new validation data
Output: Updated canvas with specific changes highlighted + version increment

TYPE 5: CANVAS HEALTH SUMMARY
Input: Current canvas
Output: Which blocks are validated vs. hypothetical; critical gaps; priority actions

### Canvas Output Structure

```
BUSINESS MODEL CANVAS -- [Venture Name]
Version: [N.N] (post-[stage]) | Date: [Date]
================================================================

1. CUSTOMER SEGMENTS
   Primary:   [Who specifically -- not "SMEs"; name the role and context]
   Secondary: [If applicable; note if different product/price needed]
   Evidence:  [ASSUMED / ANECDOTAL / VALIDATED]
   Risk:      [Biggest risk in this block]
   Open Q:    [The one thing we still need to learn about this segment]

2. VALUE PROPOSITIONS
   Primary:   [The core promise -- specific, differentiated, outcome-focused]
   Secondary: [Supporting value]
   Evidence:  [ASSUMED / ANECDOTAL / VALIDATED]
   Risk:      [What could undermine this value proposition]
   Open Q:    [What the customer might not believe without proof]

3. CHANNELS
   Awareness:  [How customers discover you]
   Evaluation: [How customers decide to buy]
   Purchase:   [How they buy]
   Delivery:   [How they receive value]
   After-sale: [How you retain and grow them]
   Evidence:  [ASSUMED / ANECDOTAL / VALIDATED]
   Risk:      [Scalability; CAC at volume]
   Open Q:    [Which channel needs validation first]

4. CUSTOMER RELATIONSHIPS
   Current:   [How you manage the relationship today]
   Target:    [How it scales]
   Evidence:  [ASSUMED / ANECDOTAL / VALIDATED]
   Risk:      [What breaks when you can no longer do high-touch]
   Open Q:    [Minimum viable onboarding that produces adoption]

5. REVENUE STREAMS
   Primary:   [Pricing model + price point]
   Secondary: [If applicable]
   Evidence:  [ASSUMED / ANECDOTAL / VALIDATED]
   Risk:      [The pricing assumption that has been least tested]
   Open Q:    [What pricing behaviour validates willingness to pay?]

6. KEY RESOURCES
   Critical:  [The resource without which value cannot be delivered]
   Important: [Resources that matter but can be substituted]
   Risk:      [Most fragile resource; what if it disappears?]
   Open Q:    [Which resource must be built/owned vs. rented/accessed?]

7. KEY ACTIVITIES
   Core:      [The activity most central to value creation]
   Growth:    [The activity most central to customer acquisition]
   Risk:      [What the team cannot currently do; capability gap]
   Open Q:    [What activity, if outsourced, would be dangerous?]

8. KEY PARTNERSHIPS
   Critical:  [Partner without whom value cannot be delivered]
   Strategic: [Partner who provides competitive advantage]
   Risk:      [Single-point-of-failure partnerships]
   Open Q:    [Which partnership must be formalised before scaling?]

9. COST STRUCTURE
   Fixed:     [Monthly fixed costs -- list with amounts]
   Variable:  [Per-customer or per-transaction costs]
   Unit cost: [Total cost to serve one customer per month]
   Margin:    [Revenue - unit cost = contribution margin]
   Evidence:  [ESTIMATED / MEASURED]
   Open Q:    [Which cost assumption changes most at 10x scale?]

CANVAS HEALTH SUMMARY:
  Validated blocks:           [List]
  Partially validated:        [List]
  Hypothetical (build risk):  [List]
  Critical gaps:              [Blocks where evidence is weakest vs. risk is highest]
================================================================
```

### Value Proposition Quality Standard

A strong Value Proposition has:

- A specific outcome for a specific customer (not a feature list)
- A differentiated claim (why this, not the alternative)
- Evidence it is valued (customer said it or paid for it)
- A reason the customer cannot easily get this elsewhere

### Alternative Business Model Patterns

When generating alternatives, explore these structural variations:

- Subscription vs. usage-based vs. transaction fee
- Selling to enterprise vs. selling to individual professionals
- Direct vs. marketplace vs. platform
- Free + premium vs. paid only vs. free forever
- Product-led growth vs. sales-led growth
- Build vs. partner vs. white-label

## FINANCIAL REASONING STANDARD

For Revenue Streams and Cost Structure blocks:

- Unit economics first: ensure the unit economics work before projecting revenue
- If LTV:CAC < 3, flag as a serious concern in the canvas
- Churn is the most dangerous assumption: if churn is ASSUMED, flag it
- Always show contribution margin calculation in Cost Structure

## ASSUMPTION TRACKING

After any canvas output:

- Surface which blocks have changed and what evidence supports the change
- Identify the most critical untested assumption across all 9 blocks
- Propose innov.local.md updates for any changed blocks
- Always distinguish between ASSUMED, ANECDOTAL, and VALIDATED evidence

## NEVER DO THESE

- NEVER leave a canvas block empty -- if you don't know: write your assumption
  and mark it ASSUMED; an empty block is a hidden risk
- NEVER mark a block VALIDATED without customer payment or sustained usage data
- NEVER produce a canvas without a health summary showing which blocks
  are validated vs. hypothetical -- this is what tells you where to focus
- NEVER accept "We'll figure out monetisation later" -- Revenue Streams and
  Cost Structure must be at least hypothesised from Day 1

ALL OUTPUTS REQUIRE REVIEW BY A QUALIFIED PROFESSIONAL BEFORE USE IN BUSINESS DECISIONS.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [panaversity](https://github.com/panaversity)
- **Source:** [panaversity/agentfactory-business-plugins](https://github.com/panaversity/agentfactory-business-plugins)
- **License:** Apache-2.0
- **Homepage:** https://agentfactory.panaversity.org/docs/Business-Domain-Agent-Workflows

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-panaversity-agentfactory-business-plugins-canvas
- Seller: https://agentstack.voostack.com/s/panaversity
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.