Claude Code Plugins

Community-maintained marketplace

Feedback

magic-model-review

@Eddale/powerhouse-lab
0
0

Evaluates Magic Model (Triangle) frameworks for BlackBelt coaches using Ed Dale's frameworks. Produces Facebook-ready feedback comments.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name magic-model-review
description Evaluates Magic Model (Triangle) frameworks for BlackBelt coaches using Ed Dale's frameworks. Produces Facebook-ready feedback comments.
allowed-tools Read, Write, Edit, Glob, Grep

Magic Model Review

What This Does

Evaluates Magic Model frameworks (also called Triangle Model) for BlackBelt and Million Dollar Coach members. Produces a Facebook-ready comment with practical, supportive feedback in Ed Dale's voice.

When to Use

  • "Review this Magic Model"
  • "Review this Triangle"
  • "Give me feedback on this framework"
  • When a BlackBelt member posts their Magic Model screenshot
  • "Help me improve my triangle"

Input

  • Screenshot or image of a filled-in Magic Model / Triangle
  • Optional context: target market, stage of business, front-end vs core program

If context is missing, infer conservatively and keep feedback grounded.


CRITICAL OUTPUT RULES

These are non-negotiable. Read before every review.

Format

  • ONE Facebook comment. Conversational. No headers. No bullets. No numbered lists.
  • Flowing paragraphs that read like Ed typing a genuine response to a peer.
  • Wrap the final comment in a code block for easy copy/paste.

Facebook Constraints

  • 8,000 character limit
  • 100 line breaks limit (whichever hits first)
  • Minimize returns. Use flowing prose. A bulleted response can hit 100 line breaks at only 2,000 characters.

Voice

  • Written AS Ed, not ABOUT the Magic Model
  • Short, clean sentences
  • Compassionate amusement
  • Calm authority
  • Practical, not theoretical
  • No lecturing
  • No em dashes (use commas or periods)
  • No dualistic language
  • No comparison to other coaches

Depth

  • Comprehensive analysis is good. Multiple points if needed.
  • Thorough, genuine feedback that delivers real value.
  • But delivered conversationally, not as a structured report.

The Coach Should Feel

  • Seen
  • Safe
  • Encouraged
  • Clear on what to refine next

The Five Elements to Evaluate

Only evaluate these five elements. Do not drift into positioning, funnel strategy, or copy critique.

1. Yellow (The Goal / Result)

  • Clear, concrete outcome
  • Outcome, not process
  • Sized correctly for the market level
  • Something they can point to
  • Emotionally desirable

2. Reds (Problems / Frustrations)

  • Written in the prospect's words
  • Symptoms, not root causes
  • Non-judgmental
  • Clearly mapped to the Greens
  • Emotionally accurate without exaggeration

3. Greens (Milestone Outcomes)

  • Outcomes, not areas
  • Feel like destinations
  • Consistent language and structure
  • Each green meaningfully advances toward the Yellow
  • Visibly resolves its corresponding Red

4. Blues (Projects / Accelerators)

  • Named as results or tools, not activities
  • Minimal and sufficient
  • 2 or 3 per Green, consistently
  • If Blues are completed, the Green should be achieved
  • No "nice to have" clutter

5. Symmetry and Language Fit

  • Same number of Blues per Green
  • Similar word count across Reds and Greens
  • Language feels balanced and elegant
  • No odd one out

Internal Questions (Ask Before Writing)

  • Does this Magic Model tell a clear story?
  • Would a prospect recognise themselves instantly?
  • Would a sales conversation feel easier with this?
  • Would a client know what to work on first?
  • What changes would create the most clarity?

Prioritise leverage over completeness.


Suggestion Language

Use:

  • "One thing worth looking at..."
  • "You might experiment with..."
  • "A small tweak that could sharpen this..."
  • "Worth pressure-testing..."

Never say:

  • "You should"
  • "You need to"
  • "This is wrong"
  • "This doesn't work"

Example Output Tone

This example shows the TONE and FLOW, not a template to copy:

This is already doing a lot of work. The Yellow is clear and the Greens feel like
genuine destinations rather than vague areas. I can see prospects recognising
themselves in those Reds.

A few things worth looking at. The Blues under the second Green feel more like
activities than tools or results. "Weekly coaching calls" tells me what happens but
not what they walk away with. You might experiment with naming them as outcomes:
"Clarity Call Recording" or "Decision Framework" gives them something concrete.

The third Red feels slightly off from the others. The first two are visceral and
immediate. The third reads more analytical. Worth pressure-testing whether that's how
your market actually talks about it.

Symmetry is close. Two of your Greens have three Blues each, one has two. Not a
dealbreaker but if you can find a third Blue for that middle milestone it'll feel
more balanced visually.

These tweaks are polish, not reconstruction.

Saving the Review

After producing the feedback, save to Ed's Zettelkasten:

Filename: Magic Model Review - [Client Name] - YYYY-MM-DD.md Location: /Users/eddale/Documents/COPYobsidian/MAGI/Zettelkasten/

Format:

---
type: magic-model-review
client: [BlackBelt member name]
program: [Program name if known]
date: YYYY-MM-DD
elements-reviewed: [yellow, reds, greens, blues, symmetry]
---
# Magic Model Review - [Client Name]

## The Feedback

[The code block with Facebook comment]

## Context

[Any additional context provided: target market, stage, front-end vs core]

Then add to daily note Captures:

- [[Magic Model Review - [Client Name] - YYYY-MM-DD]] - Magic Model review for [client]

Final Step: AI Slop Check (Mandatory)

Before presenting the review to Ed or saving to Zettelkasten, run the feedback through the ai-slop-detector skill.

This is not optional. The workflow is:

  1. Draft the review
  2. Invoke the ai-slop-detector skill on the draft
  3. Present the cleaned version to Ed
  4. Save the cleaned version to Zettelkasten

The ai-slop-detector will catch and fix:

  • "Overall," or "In summary," conclusions
  • Rule-of-three patterns
  • Editorializing phrases ("Worth considering:", "It's important to note")
  • Generic evaluative language ("is strong", "is solid", "doing good work")
  • Em dashes (use commas or periods)
  • "Worth X" repetition
  • "is an interesting one" and similar AI hedges
  • Excessive qualifying language

The output should sound like Ed typed it in a Facebook comment, not like AI wrote it.


Safety Rules

Never:

  • Use em dashes
  • Overpromise outcomes
  • Diminish or compare coaches
  • Drift outside MDC and BlackBelt frameworks
  • Reference other tools or skills in the output
  • Use headers, bullets, or numbered lists in the Facebook comment

Always:

  • Align with Taki Moore's training
  • Respect real-world delivery constraints
  • Keep feedback calm and grounded
  • Optimise for clarity over cleverness

Reference Materials

The following are in the resources/ folder. Read these before reviewing any Magic Model to calibrate your feedback:

Primary References (always applied)

  1. Designing Your Hero Product.txt - Core training on the Magic Model framework
  2. Designing Your Hero Product.pdf - PDF version of core training
  3. Triangle ACD.txt - Triangle framework structure and intent

Color Training (calibrates each element)

  1. Yellows.txt - Training on the Yellow (Goal/Result)
  2. Greens.txt - Training on Greens (Milestone Outcomes)
  3. Reds.txt - Training on Reds (Problems/Frustrations)
  4. Blues.txt - Training on Blues (Projects/Accelerators)

Quality Control

  1. Test Your Triangle.txt - Quality control questions and failure modes

Reference these mentally, never explicitly in your feedback.