# MTP (Mapping the Prompt)
> A framework for steering LLM output through sliders, grid coordinates, and presets.

Raw Markdown copies are published for AI assistants and chat tools when rendered documentation pages cannot be fetched or processed directly.

---

Source file: src/content/docs/index.md
URL: https://mappingtheprompt.com/index.md

# MTP Docs

![MTP key visual. The left panel shows node sliders for Side A and Side B, such as Open–Still and Power–Void, with Power set to 70. The right panel shows the 19x19 column:row color grid labeled A–S and 1–19.](https://mappingtheprompt.com/ogp.png)

MTP (Mapping the Prompt) is a framework for steering LLM output with grids and sliders instead of long natural-language behavior instructions. It is designed to make the ideas and concepts in a prompt easier to express intuitively, helping the user and the LLM align with fewer instructions.

Its core is a 3x3 color arrangement made of nine nodes. The relationships between color, position, polarity, and intensity are defined from this arrangement. Each color in the 3x3 arrangement defines nodes for the Side A and Side B polarities. For example, Yellow in the upper-left corresponds to Open on Side A and Still on Side B.

MTP Skill is an Agent Skill for using the MTP framework through the `/mtp` command.

## What is MTP?

Traditional prompts often adjust model behavior through natural-language instructions like these:

> Act as an expert.  
> Be more concise.  
> Think step by step.

MTP moves that behavioral steering into non-verbal parameters such as color, position, and intensity. It does not change the task itself; it adjusts qualities such as force, flow, depth, structure, openness, and focus.

## Three ways to control output

MTP Skill supports three input modes. Internally, each mode resolves into the same form: axis, polarity, and intensity.

| Mode       | Pattern                                              | Use it when                                 |
| ---------- | ---------------------------------------------------- | ------------------------------------------- |
| **Slider** | `power:100`, `flow:70`, `haze:50`                    | You need explicit controls that are easy to read. |
| **Grid**   | `J:4`, `D:16`, `A:1`                                 | You want compact steering through coordinates. |
| **Preset** | `strategist`, `synthesizer`, `maverick`, `concierge` | You want a reusable blend of multiple coordinates. |

```text
# Slider
/mtp power:100 Summarize this article

# Grid
/mtp J:4 Explain this concept

# Preset
/mtp strategist Compare these options
```

## Slider and Grid UI Preview

The MTP Interactive UI on the roadmap is planned to make sliders and grids available as visual controls. The UI is not implemented yet; at the current stage, MTP Skill is operated through the `/mtp` command.

The images below show a UI preview of sliders for controlling intensity and a grid for controlling position. They make it easier to see how coordinates such as `J:4` and `D:16` correspond to color and position. The slider view and the grid view both represent the same node system from different angles.

### Slider

![Diagram explaining MTP slider arguments using /mtp node:intensity. The left panel shows node sliders for Side A and Side B with Power set to 70. The right panel shows node intensity as 3D Chebyshev distance, with labeled directions such as Open, Power, Focus, Flow, Close, Surge, Collapse, and Fade.](https://mappingtheprompt.com/images/pages/mtp-slider-arguments-mtp-node-intensity.png)
*Slider arguments combine a node name with an intensity value, moving from the center toward a Side A or Side B direction in the MTP space.*

### Grid

![Diagram explaining MTP grid arguments using /mtp column:row. The left panel shows a simplified 3×3 color grid projected onto a 19×19 coordinate grid, with example coordinates such as J:4, D:10, P:10, and J:16. The right panel shows RGBA color values across the full 19×19 grid.](https://mappingtheprompt.com/images/pages/mtp-grid-arguments-mtp-column-row.png)
*Grid arguments select a point on the 19×19 MTP coordinate plane, where positions map back to node directions and their RGBA color values.*

In the grid UI, position is intended to be selected as if placing a point on the grid.

---

## For AI Agents

During the documentation site build, key pages are aggregated into a single `llms.txt` file. You can provide this file to AI agents to supply them with the context needed to understand and explain this site.

**For AI agents:**  
[mappingtheprompt.com/llms.txt](https://mappingtheprompt.com/llms.txt)

---

Source file: src/content/docs/foundational/grid-and-coordinate-system/index.md
URL: https://mappingtheprompt.com/foundational/grid-and-coordinate-system.md

# Grid and Coordinate System

MTP (Mapping the Prompt) is a framework for shaping prompt behavior **without** relying on explicit instructive prose, using a **dynamic model with two formal layers: Space and Intensity**. **Space** determines axis and zone from grid position; **Intensity** determines strength, polarity, and tiered constraint extraction (Volcano mapping for coordinates).

**Multiple tokens** in one `/mtp` payload (including preset expansion) add **ordered** constraint blocks: each token is still mapped through the same Space and Intensity rules; order is **not** a third parameter space or dedicated transform layer.

| Layer         | Controls                           | Key concept                                    |
| ------------- | ---------------------------------- | ---------------------------------------------- |
| **Space**     | *Where* — which axis/zone          | 19×19 grid, 3×3 macro zones, hue cycle         |
| **Intensity** | *How much* — strength and polarity | 0–100 slider, Side A / Side B duality, Volcano |

This document defines the mathematical and geometric structures that make up MTP, along with the internal parameter model used to inject metadata into prompts.

---

## Spatial structure: 19×19 grid system

The core MTP interface uses an integer coordinate system at line intersections, similar to a Go board. The grid is **19×19**.

### Symmetrical three-way division model

The space is always divided into a **geometrically balanced 3×3 macro-zone structure**. Rather than simply splitting the number of elements into three uneven groups, MTP treats the whole surface as continuous space and places grid lines and boundaries in a fully symmetrical arrangement.

**Structure from 19 grid lines**

There are 19 grid lines in total. Four of them are boundaries (outer frame and between zones); inside them lie three equal-width macro zones in sequence, each containing five grid lines.

```text
        A – F        G – L        M – S
 1–6   Yellow    |    Red      | Magenta
 7–12  Green     | Transparent | White
13–19  Cyan      |    Blue     | Purple
```

- **Line 1:** Boundary line (outer frame)
- **Lines 2–6:** Zone 1 (5 grid lines)
- **Line 7:** Boundary line (inner boundary)
- **Lines 8–12:** Zone 2 (5 grid lines)
- **Line 13:** Boundary line (inner boundary)
- **Lines 14–18:** Zone 3 (5 grid lines)
- **Line 19:** Boundary line (outer frame)

### Boundary lines and zone assignment

Boundary lines are not sharp partitions; conceptually—like a gradient—they are transition regions between macro zones. The compiler assigns each coordinate (including boundary coordinates) to **exactly one** of the nine macro-zone cells (zone bounds are defined by the compiler) and computes intensity with the same Chebyshev distance formula as for interior coordinates.

### Hue cycle and Z-order

The nine macro zones correspond to a hue cycle and also carry a generation transition order, or Z-order. Together they suggest the order in which concepts unfold and which phases processing passes through.

|                                                              | Color           | Node        | Keywords                                      |
| ------------------------------------------------------------ | --------------- | ----------- | --------------------------------------------- |
| <div class="dot-sm bg-open" aria-label="yellow"></div>       | **Yellow**      | `open`      | divergence, openness, margin                  |
| <div class="dot-sm bg-power" aria-label="red"></div>         | **Red**         | `power`     | force, assertion, rise                        |
| <div class="dot-sm bg-return" aria-label="magenta"></div>    | **Magenta**     | `return`    | return, flip, cycle                           |
| <div class="dot-sm bg-grow" aria-label="green"></div>        | **Green**       | `grow`      | growth, proliferation, layering               |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | **Transparent** | `helix`     | helix, unfold, neutral structure              |
| <div class="dot-sm bg-focus" aria-label="white"></div>       | **White**       | `focus`     | focus, convergence, precision                 |
| <div class="dot-sm bg-enter" aria-label="cyan"></div>        | **Cyan**        | `enter`     | entry, landing, structure                     |
| <div class="dot-sm bg-flow" aria-label="blue"></div>         | **Blue**        | `flow`      | flow, connection, rhythm                      |
| <div class="dot-sm bg-close" aria-label="purple"></div>      | **Purple**      | `close`     | close, completion, wrap                       |

---

## Intensity structure: three-state model and bipolarity

Each node is not a simple on/off switch, but a continuous intensity slider from 0 to 100. Polarity—positive or negative—determines whether Side A’s behavior is active or Side B’s **inverted** behavior is active.

The compiler also accepts **macro-zone color names** as slider aliases (for example, `yellow:50` ≡ `open:50`, `yellow:-50` ≡ `still:50`), using the same linear intensity mapping as the paired Side A / Side B nodes.

### Node symmetry (D4 symmetry) and pairing

Negative values activate Side B, while positive values activate Side A. Across the central baseline of `0`, they form the following dual pairs.

| Side B (-1)  | Center (0)  | Side A (+1) | Spatial Position |
| ------------ | ----------- | ----------- | ---------------- |
| `still`      | Yellow      | `open`      | Upper left       |
| `void`       | Red         | `power`     | Top              |
| `surge`      | Magenta     | `return`    | Upper right      |
| `wither`     | Green       | `grow`      | Left             |
| `collapse`   | Transparent | `helix`     | Center           |
| `haze`       | White       | `focus`     | Right            |
| `drift`      | Cyan        | `enter`     | Lower left       |
| `abyss`      | Blue        | `flow`      | Bottom           |
| `fade`       | Purple      | `close`     | Lower right      |

---

## Coordinate-to-intensity transformation model

Coordinates such as `A:1` or `J:10` on the 19×19 grid (hyphen form `A-1` is also accepted) are mapped by the MTP compiler to **polarity** (Side A / Side B) and **intensity** (0–100) using Chebyshev distance from the grid center `J:10`.

### Polarity and intensity calculation algorithm (Volcano model, Chebyshev distance)

This transformation follows a **Volcano model**: the center is neutral, the mid-ring reaches Side A peak, and the outer frame inverts into Side B. The model uses Chebyshev distance so all macro-zone centers at the same distance share the same signed value.

**Volcano model cross-section (example: row 10, from A to S):**

```text
Signed value
        Side A
 +100 |          /\                 /\
      |         /  \               /  \
      |        /    \             /    \
    0 |-------/------\-----0-----/------\-------
      |      /     Neutral Center        \
      |     /                             \
 -100 |----/                               \----
        Side B

Distance:  9      6        0        6      9
Coord:    A:10   D:10     J:10     P:10   S:10
```

Note: this is a one-dimensional cross-section (row 10). In full, the field is a 2D Chebyshev-distance ring around `J:10`—visually, a donut-shaped ring around the center.

- **Center (distance 0):** `J:10` -> signed value `0` (neutral, no constraint emitted)
- **Peak ring (distance 6):** for example, `D:10`, `D:4`, `P:16` -> Side A maximum (`+100`)
- **Outer frame (distance 9):** for example, `A:10`, `A:1`, `S:19` -> Side B maximum (`-100`)

Compiler formulas:

```text
Center (J:10)      = neutral (signed 0; no constraint)
Peak (distance 6)  = Side A maximum (signed +100)
Outer (distance 9) = Side B maximum (signed -100)

distance  = max(|x - 10|, |y - 10|)   (Chebyshev, integer 0–9)
Inner (distance ≤ 6):  signed = (distance / 6) * 100
Outer (distance > 6):  signed = 100 - 200 * (distance - 6) / 3

polarity  = +1 if signed >= 0 else -1
intensity = abs(signed), clamped to 1-100; 0 -> no constraint emitted
```

From center distance and macro-zone direction (axis), this mapping uniquely determines polarity and intensity; those values are injected into prompt metadata.

> The grid model is visualized with `scripts/mtp_grid_generator.py`. Regeneration steps and SVG details are covered in Color Grid Visualization.

---

Source file: src/content/docs/foundational/node-reference/index.md
URL: https://mappingtheprompt.com/foundational/node-reference.md

# Node Reference

## Introduction

In conventional prompt engineering, adjusting tone and behavior often relies on natural-language modifier phrasing (“act as an expert,” “think step by step”). Such phrasing tends to be ambiguous and often introduces **meta-instruction noise** alongside the task itself.

**MTP** replaces much of that control with non-verbal metadata: **coordinates and axis labels** compiled into **tiered constraints**. This page separates the **core semantics** of each Side A / Side B node from the **prompt-output tendencies** produced by the current MTP Skill.

Core semantics describe the direction represented by a node across applications. Prompt-output tendencies describe how that direction is currently translated into LLM behavior.

---

## Node quick reference

Each color axis in MTP has two nodes: Side A and Side B (each referred to as `<node>`).
Intensity (`<intensity>`) can be specified in the range `0` to `100`.

Side A is the positive-side node and can be specified by node name, such as `power:70` (`<node>:<intensity>`). Side B is the inverse-side node and can be specified by node name, such as `void:70`.

When using a color name instead of a node name, positive values activate Side A and negative values activate Side B.
For example, `red:70` activates Power, while `red:-70` activates Void.
The same polarity rule applies to node names: `power:-70` also activates the Red axis through Void.

### Side A quick reference

| | Color / Axis | Node | Usage example | Core semantics | Prompt-output tendencies |
|---|---|---|---|---|---|
| <div class="dot-sm bg-open" aria-label="yellow"></div> | Yellow | `open` | `open:100` / `yellow:100` / `D:4` | Opening, possibility, expansion | Divergent ideation, option generation, leaving room for choice |
| <div class="dot-sm bg-power" aria-label="red"></div> | Red | `power` | `power:100` / `red:100` / `J:4` | Force, assertion, drive | Firm claims, decisions, persuasive conclusions |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | Magenta | `return` | `return:100` / `magenta:100` / `P:4` | Return, reversal, recurrence, reframing | Critique, counter-reading, reframing a premise |
| <div class="dot-sm bg-grow" aria-label="green"></div> | Green | `grow` | `grow:100` / `green:100` / `D:10` | Development, layering, widening | Expanded explanations, elaboration, branching ideas |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | Transparent | `helix` | `helix:100` / `transparent:100` | Integration, unfolding structure, interrelation | Traceable reasoning, linked steps, visible structure |
| <div class="dot-sm bg-focus" aria-label="white"></div> | White | `focus` | `focus:100` / `white:100` / `P:10` | Precision, clarity, definition | Specification checks, evidence, narrowed scope |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | Cyan | `enter` | `enter:100` / `cyan:100` / `D:16` | Threshold, entry, framing | Introductions, onboarding, explicit scope and prerequisites |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | Blue | `flow` | `flow:100` / `blue:100` / `J:16` | Continuity, connection, rhythm | Smooth prose, transitions, readable progression |
| <div class="dot-sm bg-close" aria-label="purple"></div> | Purple | `close` | `close:100` / `purple:100` / `P:16` | Closure, completion, sealing | Summaries, conclusions, next actions |

### Side B quick reference

| | Color / Axis | Node | Usage example | Core semantics | Prompt-output tendencies |
|---|---|---|---|---|---|
| <div class="dot-sm bg-still" aria-label="yellow"></div> | Dark Yellow | `still` | `still:100` / `yellow:-100` / `A:1` | Restraint, preservation, suspension | Formatting, proofreading, minimal change, suppressing new suggestions |
| <div class="dot-sm bg-void" aria-label="red"></div> | Dark Red | `void` | `void:100` / `red:-100` / `J:1` | Absence, subtraction, hollowing, silence | Shortening, removing noise, dry or minimal responses |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | Dark Magenta | `surge` | `surge:100` / `magenta:-100` / `S:1` | Release, excess, eruption, overflow | High-energy writing, dense accumulation, forceful expression |
| <div class="dot-sm bg-wither" aria-label="green"></div> | Dark Green | `wither` | `wither:100` / `green:-100` / `A:10` | Attenuation, pruning, decline | Key-point summaries, concise explanations, reduced branching |
| <div class="dot-sm bg-collapse" aria-label="transparent"></div> | Translucent (Transparent) | `collapse` | `collapse:100` / `transparent:-100` | Compression, convergence, loss of structure | Direct answers, simplified summaries, flattened organization |
| <div class="dot-sm bg-haze" aria-label="white"></div> | Dark Grey (White) | `haze` | `haze:100` / `white:-100` / `S:10` | Ambiguity, diffusion, obscured definition | Poetic language, atmosphere, softened distinctions |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | Dark Cyan | `drift` | `drift:100` / `cyan:-100` / `A:19` | Deviation, wandering, loosened direction | Free association, tangents, lateral development |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | Dark Blue | `abyss` | `abyss:100` / `blue:-100` / `J:19` | Depth, descent, weight | Deep reflection, criticism, dense analysis |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | Dark Purple | `fade` | `fade:100` / `purple:-100` / `S:19` | Attenuation, dissolution, afterimage | Open-ended prose, tapering conclusions, lingering resonance |

*Side B color names are not the opposite color (White does not flip to Black); they restate the axis color to match each node's meaning. Haze is White growing hazy toward grey, shown as "Dark Grey (White)," and Collapse is Transparent breaking down, shown as "Translucent (Transparent)."*

---

## Axis structure

This section briefly organizes **how each axis relates** within the taxonomy.

- **Vertical (Red ↔ Blue):** The axis where directions that stress assertion and structure face directions that preserve receptivity and continuity.
- **Horizontal (Green ↔ White):** The axis where outward expansion while holding outline faces narrowing the subject for rigorous scrutiny.
- **Corners (Yellow, Magenta, Cyan, Purple):** Transitional **gradient** regions that connect **adjacent nodes** of the cross.
- **Center (Transparent):** A **neutral node** placed between opposing directions.

![MTP coordinate system and node layout diagram showing the unified Side A/B color and node map, plus coordinate pairs such as D:4 and A:19.](https://mappingtheprompt.com/images/pages/mtp-coordinate-system-and-node-layout.png)

*The axis layout can be read both as a 3x3 Side A / Side B map and as coordinate positions on the 19x19 grid.*

---

## Side A / Side B

As an output tendency, Side A often appears as a force that moves the response forward. In an introduction, for example, it may invite the reader in, add reasons, structure the material, sharpen focus, or carry the text toward a conclusion.

Side B is not simply a weaker Side A. As the inverted pole of the same axis, it often shifts output away from construction toward reduction, away from clarity toward ambiguity, away from expansion toward descent, and away from closure toward afterimage. Side B is not the "bad" or "low" side; it functions as the shadow or reverse face of the axis.

---

## Side B interpretation

Under the **Chebyshev-based radial rule**, coordinates on the **outer perimeter frame** flip to **Side B** for that zone. In constraint design, Side B is expressed as an **inverted pole** of the same axis: a **paired opposite** in a yin-yang sense (for example, Power → Void, Focus → Haze, Grow → Wither).

Side B suits cases that call for a **stronger extreme** along that axis or **more compressed** surface wording, with the same model-to-model variability as Side A.

---

## Combining nodes and presets

MTP can also specify multiple nodes at the same time to blend output tendencies. Instead of switching output with a single node, it can layer several axes like a gradient, such as combining assertion (`power:20`) with precision (`focus:30`), or development (`grow:20`) with readability (`flow:50`).

The idea is to reduce the burden of operation by combining MTP nodes intuitively, while still changing the model's output tendency in a controlled way.

When several MTP tokens are present, **each token** is still resolved under the same **Space / Intensity** rules; the compiler emits one constraint block **per token in parse order** (the order they appear in the payload).

That order is **not** a third parameter space or a dedicated “motion” layer. The sequence *may* be read as a **semantic trajectory** on the grid, but that is an **optional design reading** only.

**Named sliders** (`power:100`, `flow:70`) are an easy starting point when meaning should stay readable; **grid coordinates** suit cases where the visible message should remain short.

**Named presets** expand to fixed coordinate sequences defined in `skills/mtp/references/presets.yaml`; they can be applied to reproducible multi-step blends.

---

Source file: src/content/docs/foundational/overview/index.md
URL: https://mappingtheprompt.com/foundational/overview.md

# Overview

MTP (Mapping the Prompt) is a framework for steering LLM output through structured controls — sliders, grid coordinates, and presets — instead of long natural-language behavior instructions.

It controls model tone and style through a **coordinate-based compilation step** rather than direct verbal instructions. Instead of relying on requests such as "be concise" or "think deeply," it uses a position in a structured space: a slider value, a grid coordinate, or a named preset. The compiler deterministically converts that position into constraint XML that reshapes the model's output.

This page summarizes **core terms** and a lightweight description of the two layers behind MTP's controls. In practical terms, those layers are **grid position** and **intensity / polarity**; in the formal model they are called **Space** and **Intensity**. Full equations and zone boundaries are covered in Grid and Coordinate System; per-node behavior is summarized in Node Reference.

The 3x3 color arrangement used by Space is:

| Position | Left        | Center        | Right     |
| -------- | ----------- | ------------- | --------- |
| Top      | Yellow      | Red           | Magenta   |
| Middle   | Green       | Transparent   | White     |
| Bottom   | Cyan        | Blue          | Purple    |

---

## Core terms

The following terms recur throughout MTP documentation. Together they describe the pipeline from user input to applied constraints:

```text
  ┌─────────┐     ┌──────────────────┐     ┌───────────────────────────┐     ┌─────────────┐
  │  Input  │ ──→ │     Compiler     │ ──→ │  Constraint extraction    │ ──→ │   Output    │
  └─────────┘     └──────────────────┘     └───────────────────────────┘     └─────────────┘

  preset ─── expands to ──→ grid coordinates ──┐
  slider (node:intensity) ───────────────────→ ├→ (axis, polarity, intensity) → node file → constraints
  grid coordinate (column:row) ──────────────→ ┘
```

| Term                | Meaning                                                                                                                                           |
| ------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Node**            | One of the nine semantic axes (for example, Red). Each node has a Side A name (for example, Power) and a Side B name (for example, Void).         |
| **Color / Axis**            | The color-based identity used to distinguish nodes: Yellow, Red, Magenta, Green, Transparent, White, Cyan, Blue, Purple.                          |
| **Side A / Side B** | The two poles of each axis. Positive intensity activates Side A; negative intensity activates Side B.                                             |
| **Intensity**       | A strength value from 0 to 100. It determines which constraint tier (Low / Mid / High) is extracted.                                              |
| **Slider**          | Input in `node:intensity` format. Examples: `power:70`, `void:80`, `yellow:-30`                                                                   |
| **Grid** | Input in `column:row` format on the 19×19 grid. Examples: `D:16`, `A:1`. The compiler calculates axis, polarity, and intensity from the position. |
| **Preset**          | A named blend that expands into grid coordinates before parsing. Example: `strategist` expands to `P:16 P:4`.                                     |

All input paths ultimately converge on the same internal representation: **(axis, polarity, intensity)**. Inputs such as `power:100`, `red:100`, and `J:4` resolve to the same three components and therefore extract the same constraints.

| | Color / Axis | Side A Node | Side B Node   |
| --- | ----------- | ------ | -------- |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | Yellow      | `open`   | `still`    |
| <div class="dot-sm bg-power" aria-label="red"></div> | Red         | `power`  | `void`     |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | Magenta     | `return` | `surge`    |
| <div class="dot-sm bg-grow" aria-label="green"></div> | Green       | `grow`   | `wither`   |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | Transparent | `helix`  | `collapse` |
| <div class="dot-sm bg-focus" aria-label="white"></div> | White       | `focus`  | `haze`     |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | Cyan        | `enter`  | `drift`    |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | Blue        | `flow`   | `abyss`     |
| <div class="dot-sm bg-close" aria-label="purple"></div> | Purple      | `close`  | `fade`     |

---

Source file: src/content/docs/optional/design-background/index.md
URL: https://mappingtheprompt.com/optional/design-background.md

# Design Background

> [!NOTE]
> This document is not required reading. It is optional reference material for understanding the design rationale.
>
> The UI images on this page are conceptual diagrams of the design. The current MTP skill is text-based and is used through `/mtp <args>` sliders, grid coordinates, and presets.

**MTP (Mapping the Prompt)** maps intent to a discrete color grid instead of relying on long verbal prompt instructions. This document explains the design rationale for that grid: why the layout is 3×3, why **Transparent** sits at the center, and how the surrounding colors relate to both **Yin–Yang and Five-Elements thought** and a **hue cycle**.

MTP color nodes are not labels for assigning a fixed role or persona to an AI. They are control units that arrange multiple output tendencies as directions and steer the response by combining direction and intensity.

---

## 3×3 grid

The core of MTP is a 3×3 grid of nine cells (nodes). Each cell carries both a color identity and a semantic role.

```text
+-----------+-------------+-------------+-------------+
| Position  | Left        | Center      | Right       |
+-----------+-------------+-------------+-------------+
| Top       | YELLOW      | RED         | MAGENTA     |
+-----------+-------------+-------------+-------------+
| Middle    | GREEN       | TRANSPARENT | WHITE       |
+-----------+-------------+-------------+-------------+
| Bottom    | CYAN        | BLUE        | PURPLE      |
+-----------+-------------+-------------+-------------+
```

In implementation, this macro layout expands into a **19×19** coordinate grid. Behind this arrangement lies a distinctive design that overlays two organizing principles:

- A directional reading drawn from **Wu Xing (Five Elements)**.
- Corner placement along a **discrete hue cycle**.

Wu Xing directions and hue-cycle corners are combined in one layout. This is **not** offered as the authoritative arrangement for cultural or intellectual history; it is explained as a **mnemonic for which semantic direction each node faces**.

**Design Visualization**
![A prism-centered conceptual color map with eight color nodes around it, paired with a Wu Xing directional diagram placing Wood, Fire, Earth, Metal, and Water at left, top, center, right, and bottom.](https://mappingtheprompt.com/images/pages/mtp-design-background-from-wu-xing-to-color-nodes.png)
*MTP combines cues from the light spectrum and hue cycle with a Wu Xing-inspired directional model to organize the semantic direction of its color nodes.*

---

## Why MTP's center is Transparent

The key to the center is the role of **Earth (土)** within Wu Xing — in the Five-Elements palette often associated with yellow or brown. Wood, Fire, Metal, and Water occupy directional positions; Earth occupies the **center**. Here, Earth is not read as another outward-moving element, but as a **medium** or **field** that makes transition and balance possible.

```text
                  +-----------------+
                  |   Fire (South)  |
+-----------------+-----------------+-----------------+
|   Wood (East)   | Earth (Center)  |  Metal (West)   |
+-----------------+-----------------+-----------------+
                  |  Water (North)  |
                  +-----------------+
```

### “Earth” as spatial foundation

Relative to the four directions, the place from which orientation is taken is the center. In that sense, “Earth” is the ground on which the other phases unfold. The directional phases can oppose, generate, or transform one another, but the center remains the point that holds the system together.

### “Earth” as seasonal converter

Wu Xing is often read through the seasons: Wood with spring, Fire with summer, Metal with autumn, and Water with winter. Earth does not claim an exclusive season of its own in the same way. Instead, it is associated with **Doyo (土用)**, the transitional interval between seasons. Under that reading, Earth is the phase through which one season returns and another emerges. It is less a fixed season than a **converter between seasons**.

### “Earth” as balancer

Wood and Fire are commonly read as more Yang-oriented phases, while Metal and Water are more Yin-oriented. Earth occupies the balancing role between those poles. A direct jump from one extreme to another suggests rupture; Earth receives that change, steadies it, transforms it, and makes continuation possible.

### “Earth” as field and Transparent

In MTP, the center (Earth) is **Transparent**, not yellow. Wood, Fire, Metal, and Water can be treated as directional vectors. Earth, in this reading, is different: it is the **field or substrate that receives, supports, and mediates** those vectors.

MTP encodes that role as **Transparent (Helix)** rather than as a fifth hue. The point is not “another color in the middle,” but a mediating center that stays off the hue poles while remaining involved in every direction.

### The 4+1 pattern across world traditions

The layout also connects to a broader **4+1** motif found in multiple traditions: four directional poles plus one mediating center. Greek discussions of **Aether** and Indian discussions of **Akasha** provide loose analogies. In MTP, however, the main anchor remains the Wu Xing reading of **Earth-as-medium**. Cross-cultural comparison is a teaching aid, not evidence. The layout is **not** intended as a mystical reading; it points to a shared intuition about nature and placement — for example, fire rises and water seeks the low — rather than serving as proof of a single esoteric schema.

For example, Red and Blue form a strong vertical opposition, Green and White form a horizontal complement, and Transparent mediates between them without becoming a directional pole itself.

---

## From Five Elements to nine colors

Wu Xing provides five named phases, while the grid has nine cells. The difference is handled by using the Five Elements as the **cross skeleton** and assigning the four corners to **transitional positions** between adjacent phases. Those transitions are then aligned with intermediate hues on a standard hue wheel.

### Cross skeleton

| Direction | Five Elements | Grid position | MTP color |
| --- | --- | --- | --- |
| South | Fire | Top center | **Red** |
| East | Wood | Middle left | **Green** |
| Center | Earth | Center | **Transparent** |
| West | Metal | Middle right | **White** |
| North | Water | Bottom center | **Blue** |

The Metal → **White** correspondence reflects a design association of brightness and reflectivity. It is a metaphor within the grid design, not a physics claim.

### Corners as transitions

| Position | Adjacent elements | Seasonal transition | MTP color |
| --- | --- | --- | --- |
| Top left | Wood(Green) ↔ Fire(Red) | Spring → Summer | **Yellow** |
| Top right | Fire(Red) ↔ Metal(White) | Summer → Autumn | **Magenta** |
| Bottom left | Water(Blue) ↔ Wood(Green) | Winter → Spring | **Cyan** |
| Bottom right | Metal(White) ↔ Water(Blue) | Autumn → Winter | **Purple** |

The corners are not merely blended values between two adjacent directions. Yellow, Magenta, Cyan, and Purple are each defined as independent Markdown nodes with their own semantic directions at those transitional positions.

On HSV/HSL, Yellow sits between Green and Red (roughly 60°), and Cyan between Green and Blue (roughly 180°). That is close enough to standard hue order to give the corner placements structural coherence. Again, the point is design alignment, not a rigorous derivation from color science.

---

## Structural properties of the grid

The same design rationale can also be read structurally: the nine-node color layout forms both a conceptual map and a directional control surface.

![An interactive control diagram with eight Markdown color nodes around an origin, where drag direction and distance indicate node selection and intensity, paired with a four-direction Red, Green, White, and Blue layout.](https://mappingtheprompt.com/images/pages/mtp-interface-background-from-color-nodes-to-directional-control.png)
*Each color node is defined as a Markdown instruction; direction from the center selects a node, while distance sets its intensity.*

### Direction and intensity as controls

Like a joystick centered on an origin, MTP adjusts the qualities applied to LLM output according to the direction and distance of movement. Each color node is defined as a `.md` instruction, so selecting a color is equivalent to selecting the corresponding instruction module.

### From the 3×3 macro layout to 19×19 coordinate space

The 3×3 grid is the semantic skeleton of the nine node directions. In implementation, MTP expands this macro layout into a 19×19 coordinate space and resolves axis, polarity, and intensity from a coordinate. This makes it possible to specify positions between nodes rather than selecting only discrete nodes.

![Nine Markdown nodes arranged in 3×3 macro zones, paired with an interpolated 19×19 color grid highlighting Power at J:4 and Grow at D:10.](https://mappingtheprompt.com/images/pages/mtp-coordinate-space-from-color-nodes-to-19x19-control.png)
*The 3×3 grid provides the semantic skeleton, while the 19×19 grid provides a coordinate surface for finer control of direction, polarity, and intensity.*

The coordinate model for resolving polarity and intensity is documented in [Grid and Coordinate System](https://mappingtheprompt.com/foundational/grid-and-coordinate-system/).

### Lattice embedding of the hue cycle

The hue cycle is normally continuous. MTP **discretely embeds** it into a 3×3 lattice as eight peripheral hues around a non-hue center. This preserves circulation around the outside while allowing the grid to function as a navigable coordinate surface.

### Transparent as a medium node

Many color systems place an achromatic value such as gray, white, or black at the center. MTP chooses **Transparent** instead. That decision treats the center not as a neutral color sample, but as a **medium-like node**: colorless in itself, yet involved in the transmission and integration of surrounding directions.

### Z-order and semantic movement

Scanning the 3×3 grid from top-left toward bottom-right in a row-wise zigzag (Z-order) yields this color sequence:

```text
Yellow → Red → Magenta → Green → Transparent → White → Cyan → Blue → Purple
```

This ordering has **structural significance** within the macro layout.

- **Warm-to-cool tendency by row:** Roughly, the top row reads warmer, the middle row more neutral / mediating, and the bottom row cooler. This is a loose visual reading, not a strict color-temperature or scientific claim.
- **The middle row of three:** **Green** (chromatic) → **Transparent** (medium) → **White** (achromatic leaning) sit in one row. This can be loosely related to how hue and lightness axes intersect in classical **opponent-process** accounts — as an **analogy** for looking at the design, not a derivation from color science.

In image processing, Z-order is often just a traversal sequence chosen for computational efficiency. In MTP, the same ordering can also be read as a **color trajectory** through the space. It can help when thinking about **semantic trajectory** across ordered tokens or presets.

---

Source file: src/content/docs/optional/mapping-and-sequence/index.md
URL: https://mappingtheprompt.com/optional/mapping-and-sequence.md

# Mapping and Sequence

> [!NOTE]
> This document is supplementary material. It is intended to support understanding of the design background.

MTP can be read not only as a tool for prompt steering, but also as a framework for **mapping concepts onto a 3×3 grid** and as an **ordered sequence** of node poles. That follows from how MTP node placement was designed using widely shared cues from color and direction.

This document covers two extended readings:

1. **Mapping on the grid** — placing characters, archetypes, prompt tendencies, output tendencies, and similar notions onto MTP nodes.
2. **Sequence (Z-order)** — taking input and output as a frame and reading the 18 node poles between them as a linear arc.

---

## Concept mapping on the grid

The MTP grid can be used as a surface for mapping other conceptual systems.
For example, the following kinds of alignment are possible.

### Reference: base grid

```text
3×3 color map (macro zones):
+-------------+-------------+-------------+
| Yellow      | Red         | Magenta     |
+-------------+-------------+-------------+
| Green       | Transparent | White       |
+-------------+-------------+-------------+
| Cyan        | Blue        | Purple      |
+-------------+-------------+-------------+

Node map (Side A):
+-------------+-------------+-------------+
| Open        | Power       | Return      |
+-------------+-------------+-------------+
| Grow        | Helix       | Focus       |
+-------------+-------------+-------------+
| Enter       | Flow        | Close       |
+-------------+-------------+-------------+

Node map (Side B):
+-------------+-------------+-------------+
| Still       | Void        | Surge       |
+-------------+-------------+-------------+
| Wither      | Collapse    | Haze        |
+-------------+-------------+-------------+
| Drift       | Abyss       | Fade        |
+-------------+-------------+-------------+
```

### Example: character mapping

A character system can be placed on the grid in an intuitive way.
As one example, Pixar’s [*Inside Out*](https://en.wikipedia.org/wiki/Inside_Out_(2015_film)) is built around emotional functions, so it can be read against the MTP grid.

**Side A** (*Inside Out*):

```text
+-------------+-------------+-------------+
| Joy         | Anger       | Joy         |
+-------------+-------------+-------------+
| Disgust     | Riley       | Fear        |
+-------------+-------------+-------------+
| Joy         | Sadness     | Joy         |
+-------------+-------------+-------------+
```

Joy is placed on Open, Return, Enter, and Close as an integrative motion around Riley’s experience. In MTP, the four corners can be read as threshold movements: opening possibility, turning experience back into meaning, entering situations, and bringing episodes to completion. In Inside Out, Joy keeps Riley’s inner world narratively continuous and emotionally livable, shifting roles as needed. This makes the mapping a philosophical reading of Joy as Side A motion, rather than a literal trait match.

Example alignment:

| | Node | Character | Character reading |
| --- | --- | --- | --- |
| <div class="dot-group" role="img" aria-label="Open, Return, Enter, and Close: the four corner nodes on the MTP grid"><div class="dot-sm bg-open" aria-hidden="true"></div><div class="dot-sm bg-return" aria-hidden="true"></div><div class="dot-sm bg-enter" aria-hidden="true"></div><div class="dot-sm bg-close" aria-hidden="true"></div></div> | `open`, `return`, `enter`, `close` | Joy       | Joy opens possibility, reframes experience with hope, and carries it toward bright completion. |
| <div class="dot-sm bg-power" aria-label="red"></div>         | `power` | Anger     | Anger and assertion drive impulsive but forceful action. |
| <div class="dot-sm bg-grow" aria-label="green"></div>        | `grow`  | Disgust   | Disgust and sorting tune pleasant–unpleasant boundaries and self-defense. |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | `helix` | Riley     | Many emotions pass through; the person’s neutral hub. |
| <div class="dot-sm bg-focus" aria-label="white"></div>       | `focus` | Fear      | Anxiety and vigilance turn attention toward danger and uncertainty. |
| <div class="dot-sm bg-flow" aria-label="blue"></div>         | `flow`  | Sadness   | Sadness and acceptance deepen emotional flow and meaning. |

**Side B** (*Inside Out 2*):

```text
+------------------+------------------------+------------------+
| Bloofy           | Lance Slashblade       | Anxiety          |
+------------------+------------------------+------------------+
| Embarrassment    | Riley                  | Ennui            |
+------------------+------------------------+------------------+
| Envy             | Deep Dark Secret       | Nostalgia        |
+------------------+------------------------+------------------+
```

Example alignment:

| | Node | Character | Character reading |
| --- | --- | --- | --- |
| <div class="dot-sm bg-still" aria-label="yellow"></div>         | `still`    | Bloofy             | A childhood imaginary comfort character stills real motion into consolation. |
| <div class="dot-sm bg-void" aria-label="red"></div>             | `void`     | Lance Slashblade   | Heroic fantasy hardens into empty, unusable force. |
| <div class="dot-sm bg-surge" aria-label="magenta"></div>        | `surge`    | Anxiety            | Anxiety peaks, heightening thought and bodily response excessively. |
| <div class="dot-sm bg-wither" aria-label="green"></div>         | `wither`   | Embarrassment      | Shame weakens boundary defense, shrinking the self under self-consciousness. |
| <div class="dot-sm bg-collapse" aria-label="transparent"></div> | `collapse` | Riley              | The neutral hub overloads and collapses; emotional integration briefly fails. |
| <div class="dot-sm bg-haze" aria-label="white"></div>           | `haze`     | Ennui              | Indifference and irony diffuse vigilance and cloud attention. |
| <div class="dot-sm bg-drift" aria-label="cyan"></div>           | `drift`    | Envy               | Craving loses direction, repeating insatiable want for what others have. |
| <div class="dot-sm bg-abyss" aria-label="blue"></div>           | `abyss`    | Deep Dark Secret   | Buried secrets sink, pulling self-understanding toward the bottom. |
| <div class="dot-sm bg-fade" aria-label="purple"></div>          | `fade`     | Nostalgia          | After closure, faint regret lingers, turning toward longing for the past. |

Each cell can also be read as the **inverted pole** of the Side A node at the same coordinates (for example, Open’s optimism becomes Still’s arrest; Power’s thrust becomes Void’s hollowing-out).

---

## Ordered sequence reading

MTP is formally defined by grid position, polarity, and intensity, but the nodes can also be read as a sequence.
Tracing the 3×3 grid from top-left to bottom-right in a row-wise zigzag (Z-order) yields this color order:

```text
Yellow → Red → Magenta → Green → Transparent → White → Cyan → Blue → Purple
```

### 20-node sequence: 1+9+9+1

Replacing colors with nodes, add **Start (Input)** and **End (Output)** to the **nine Side A nodes** and **nine Side B nodes** to form a linear arc:

```text
1. Start (Input)          ← beginning (input)
2. Open                   ┐
3. Power                  │
4. Return                 │
5. Grow                   │
6. Helix                  │  Side A: positive pole
7. Focus                  │
8. Enter                  │
9. Flow                   │
10. Close                  ┘
11. Still                  ┐
12. Void                   │
13. Surge                  │
14. Wither                 │
15. Collapse               │  Side B: negative pole
16. Haze                   │
17. Drift                  │
18. Abyss                  │
19. Fade                   ┘
20. End (Output)           ← end (output)
```

In this reading, the order follows the grid Z-order explained in Design Background.
**Start** and **End** are not nodes on the grid; they are virtual frame nodes added for sequence reading, marking the start and end of the arc.
The eighteen Side A and Side B nodes form their respective semantic frames.
As plain text, that is a **1+9+9+1** structure:

- **1**: Start (Input)
- **9**: Side A nodes
- **9**: Side B nodes
- **1**: End (Output)

Overall, the sequence forms a two-part arc:

- **Side A**: expansion, energy, unfoldment, arrival
- **Side B**: inversion, depletion, dismantling, attenuation

---

## Generality of classification and order

These readings help when:

- building an intuitive picture of **MTP Skill** from character placement
- mapping a 20-track AI-chosen music playlist onto the 20-node sequence

---

Source file: src/content/docs/optional/color-grid-visualization/index.md
URL: https://mappingtheprompt.com/optional/color-grid-visualization.md

# Color Grid Visualization

The script described in this document outputs an SVG of MTP’s **color grid**. It visualizes the spatial and intensity models explained in Grid and Coordinate System, and is intended as a basis for a planned interactive generative UI.

## Grid variants

Several SVGs are checked in under `public/images/grids/`. The `package.json` `scripts` map includes `build:grid`, `build:grid-10x10`, `build:grid-28x28`, and `build:grid-37x37` so each asset can be regenerated from `scripts/mtp_grid_generator.py` without hand-editing paths.

File names use **line counts** per side (grid intersections): an *N*×*N* cell checkerboard is bounded by (*N*+1)×(*N*+1) lines, so the default 18×18 cells → `mtp-grid-19x19.svg`.

| Preview | Lines (per side) | Cells | Path |
| :---: | :--- | :--- | :--- |
| <img src="../../images/grids/mtp-grid-37x37.svg" alt="MTP color grid, 37×37 lines (36×36 cells)" width="200"> | 37×37 | 36×36 | `public/images/grids/mtp-grid-37x37.svg` (`build:grid-37x37`) |
| <img src="../../images/grids/mtp-grid-28x28.svg" alt="MTP color grid, 28×28 lines (27×27 cells)" width="200"> | 28×28 | 27×27 | `public/images/grids/mtp-grid-28x28.svg` (`build:grid-28x28`) |
| <img src="../../images/grids/mtp-grid-19x19.svg" alt="MTP color grid, 19×19 lines (18×18 cells)" width="200"> | 19×19 | 18×18 | `public/images/grids/mtp-grid-19x19.svg` (`build:grid`) |
| <img src="../../images/grids/mtp-grid-10x10.svg" alt="MTP color grid, 10×10 lines (9×9 cells)" width="200"> | 10×10 | 9×9 | `public/images/grids/mtp-grid-10x10.svg` (`build:grid-10x10`) |

## How it works

The generator uses one spatial model and one palette set; only the grid size (and pixel cell size) changes between presets. The default asset is an **18×18 cell** grid (the cell regions bounded by **19×19** line intersections).

- **Cell color** comes from the **hue cycle** — for each cell, it is determined by the angle from the center.
- **Opacity / brightness** comes from **Chebyshev distance** (Volcano model). Intensity peaks at distance 6 and falls off toward the center and the outer frame.
- Under the color grid, a **checkerboard layer** indicates transparency. The background rectangle, checkerboard colors, and palette pairs are defined at the top of the script.

Strict specification of primary or hue-wheel colors is not required, but the script is designed to respect the grid’s color relationships. The script defines an `inner_palette` and `outer_palette`: these are the inner and outer anchor colors within a single SVG.

## Generating the SVG

From the repository root:

```bash
npm run build:grid
npm run build:grids
```

The `npm` scripts write directly to their matching files under `public/images/grids/`. To emit raw SVG to standard output instead, call the Python generator directly:

```bash
python3 scripts/mtp_grid_generator.py --grid-28x28
python3 scripts/mtp_grid_generator.py --grid-37x37
```

---

## MTP Interactive UI preview

In the planned **MTP Interactive UI**, the generated 19×19 SVG color grid serves as the central coordinate surface. The images below are UI previews with the grid centered and coordinate labels around the frame; they are included for now to verify how grid coordinates map to colors.

| Variant A | Variant B | Variant C |
| :---: | :---: | :---: |
| <img src="../../images/pages/mtp-ui-grid-guide--a.png" alt="MTP Interactive UI preview, variant A" width="320" loading="lazy" decoding="async"> | <img src="../../images/pages/mtp-ui-grid-guide--b.png" alt="MTP Interactive UI preview, variant B" width="320" loading="lazy" decoding="async"> | <img src="../../images/pages/mtp-ui-grid-guide--c.png" alt="MTP Interactive UI preview, variant C" width="320" loading="lazy" decoding="async"> |

The image at the center of UI preview A is the default `mtp-grid-19x19.svg` output from `scripts/mtp_grid_generator.py`. Column labels `A`–`S` and row labels `1`–`19` are added by the UI frame, so the same layout can also be used as a reference sheet for checking `/mtp <column:row>` positions.

For example, `J:10` marks the neutral center, while coordinates such as `J:4`, `D:16`, or `P:16` can be read from the labeled grid before being passed to `/mtp`.

---

Source file: src/content/docs/skills/index.md
URL: https://mappingtheprompt.com/skills.md

# Skills Installation

MTP offers several Agent Skills, and this page covers installation for all of them. The same steps apply to [MTP Skill](https://mappingtheprompt.com/skills/mtp/) and [MTP Playlist Skill](https://mappingtheprompt.com/skills/mtp-playlist/). Choose the method supported by the host where the skill will run.

## ZIP

Download the skill you want to add.

- [Download `mtp-skill.zip`](https://mappingtheprompt.com/downloads/mtp-skill.zip)
- [Download `mtp-playlist-skill.zip`](https://mappingtheprompt.com/downloads/mtp-playlist-skill.zip)

Each ZIP contains the Skill package built from its directory under `skills/` in the repository.

### Hosts that support ZIP upload

A growing number of hosts support uploaded custom Agent Skills, including Claude, Manus, Grok, and ChatGPT. The exact labels and entry points differ by host and may change, so check the host's official help when installing.

Basic flow:

1. Download the ZIP for the skill you want.
2. Open the host's custom skill settings, such as `Customize > Skills`.
3. Add a custom skill and upload the ZIP file.
4. Confirm that the skill appears in the skill list and is enabled.
5. In chat, run a prompt that invokes the skill.

[Use skills in Claude](https://support.claude.com/en/articles/12512180-use-skills-in-claude) ↗

ZIP upload is usually performed in a desktop or web client. After the skill is added and enabled for the account, mobile availability depends on the host's skill sync behavior and supported clients.

## CLI

### GitHub CLI (`gh`)

The GitHub CLI `gh skill` command is a preview feature and may change.

```bash
gh skill install imkohenauser/mtp skills/mtp
```

```bash
gh skill install imkohenauser/mtp skills/mtp-playlist
```

[GitHub CLI `gh skill` documentation](https://cli.github.com/manual/gh_skill) ↗

### Vercel Skills CLI (`npx skills`)

```bash
npx skills add imkohenauser/mtp --skill mtp
```

```bash
npx skills add imkohenauser/mtp --skill mtp-playlist
```

[Vercel Skills CLI documentation](https://www.skills.sh/docs/cli) ↗

## Check

After adding a skill, run a prompt that invokes it.

```text
/mtp power:100
Compared with other major AI models from competing companies, please explain your strengths.
```

```text
/mtp-playlist Madonna from present to past
```

## Next

- [MTP Skill](https://mappingtheprompt.com/skills/mtp/)
- [MTP Playlist Skill](https://mappingtheprompt.com/skills/mtp-playlist/)

---

Source file: src/content/docs/skills/mtp-playlist/index.md
URL: https://mappingtheprompt.com/skills/mtp-playlist.md

# MTP Playlist Skill

MTP Playlist is an **Agent Skill** that builds music sequences with the Mapping the Prompt (MTP) framework.

It is not just a mood-based track list.  
It maps each track to an MTP node and shapes a playlist that can explain "why this track," "why this node," and "why this order."  
It outputs a readable playlist deliverable in Markdown.

- [Download the MTP Playlist Skill zip](https://mappingtheprompt.com/downloads/mtp-playlist-skill.zip)
- [Installation guide](https://mappingtheprompt.com/skills/)

---

## Read tracks as placement

MTP Playlist treats a playlist like a card deck or a tarot spread.  
The left side is Side A (Start and nine nodes); the right is Side B (nine nodes and End). Each track has a distinct role at its position.

**Start** and **End** are not nodes on the grid; they are virtual frame nodes (conceptual slots) added for sequence reading.  
In MTP Playlist, they are handled as semi-transparent glass nodes that float, shifting toward surrounding nodes according to the assigned track's character.  
For example, if the track at Start is forceful, it drifts toward the Power direction on the Red axis.

Placement itself builds the playlist's meaning.

![Structure diagram for MTP Playlist. Side A on the left and Side B on the right. Each cell shows a record-like circle on a color-coded background; Start and End appear as semi-transparent glass nodes floating above the layout.](https://mappingtheprompt.com/images/pages/mtp-playlist-reading-songs-as-placement.png)

*`1+9+9+1` Side A (left) and Side B (right).*

---

## Basic usage

Call the skill explicitly by writing a theme after `/mtp-playlist`.

### Specify an artist, track, or theme

Set the direction of the playlist with a genre, a group or label, an era, a scene, or any keyword.

```text
/mtp-playlist Morning 6AM
/mtp-playlist Mozart for Spring, 60-tracks
/mtp-playlist Madness in Classical Music
/mtp-playlist Mephistophelean Classics
/mtp-playlist Modern Jazz with Vibraphone
/mtp-playlist Melancholic Bossa Nova and Saudade
/mtp-playlist Movie Drive Songs
/mtp-playlist Moonlit 80s Heavy Metal
/mtp-playlist Motown Deep Cuts
/mtp-playlist Michael Jackson, Minor Songs
/mtp-playlist Mic Relay 1970s–1980s Rap
/mtp-playlist Marley’s Rock
/mtp-playlist MDR, Berghain Techno
/mtp-playlist Marilyn Manson Side B, David Bowie Side A
/mtp-playlist Madonna 2026
```

### Anchor a specific track to a specific node

This is a way to pin a specific track to an MTP node and build the whole flow from there.  
Placing a favorite or signature track at Start, Helix, Abyss, Collapse, or another node changes the surrounding selections and the overall impression, even for the same track.

```text
/mtp-playlist Marvin Gaye "Inner City Blues" at Abyss, 70s Soul
/mtp-playlist MJB "Real Love" at Helix, 90s R&B
/mtp-playlist Mobb Deep "Shook Ones Pt. II" at Start, 90s New York Rap
/mtp-playlist Missy Elliott "Get Ur Freak On" at Start, Y2K Hip-Hop
/mtp-playlist Merry Christmas Mr. Lawrence at Collapse, Piano
```

---

## Playlist format

By default, it builds a 20-track sequence in a `1+9+9+1` structure.

* `#1 Start`
* `#2–10` Side A
* `#11–19` Side B
* `#20 End`

You can request other sizes explicitly: a 10-track Side A or Side B half, 30 tracks, 40 tracks, 60 tracks, or a shorter mapping when the source has fewer tracks.

For this ordering model, see [Mapping and sequence](https://mappingtheprompt.com/optional/mapping-and-sequence/).

---

## What you get

This skill produces a Markdown **playlist deliverable**.  
On its own, the skill does not create a playable playlist.

In ChatGPT with a linked Apple Music account, you can build an Apple Music playlist from the deliverable's track list.  
In that case, you can hand over the "copy-friendly track list" included in the deliverable as is.

It contains the following:

* A title and a short description
* The original prompt
* Playlist metadata
* A copy-friendly track list
* A track-to-node mapping table
* The selection rationale and notes

Each line explains how the assigned track makes its node work.  
The explanation rests on what you can actually hear: sound, rhythm, arrangement, vocals, lyrics, production, historical role, function within a scene, and cultural effect.

---

## What MTP Playlist addresses

A playlist can end up as a collection of tracks while rarely becoming a single experience that has an order.  
MTP Playlist addresses that gap through node placement and explanation.

### The intent behind the order

In many playlists, neither the selection rationale nor the intent behind the order is visible.

Because MTP Playlist gives each track a position and a reason, you can follow the order and its intent not only by listening but also by reading.

### Matching conditions versus a musical experience

AI-generated playlists are good at gathering tracks that match conditions such as genre, mood, era, activity, or artist.  
But gathering tracks that match conditions is not the same as building a musical experience that has an order.

MTP Playlist does more than pick tracks: it decides what role each track plays within the sequence.

### Balancing the familiar and the discovery

A playlist that leans too far toward standards feels overly familiar, while adding too many discoveries makes the flow easy to break.

MTP Playlist weighs a track's role at its position over its popularity.  
Well-known tracks are used as anchors or centers.  
Deep cuts and peripheral tracks are also chosen when the position calls for them.

### How it differs from shuffle

Shuffle can break the familiarity of an order.  
But it also destroys the meaning of the order itself.

MTP Playlist does not create freshness by randomizing the order.  
It creates a new reading by placing different tracks within the same structure.

### Interpretation per model

MTP Playlist is not a skill for making every model produce the same correct answer.

For the same theme, the selection changes with the model or the curator.  
A lightweight model has less time to explore and keeps less context, so it cannot fully review the fit with the surrounding tracks and tends toward a narrower selection.

Because of the training-data cutoff, the latest releases and trends may not be reflected.  
Some outputs follow the constraints strictly, while others emphasize narrative or lean on a greater share of signature tracks.

This difference is close to the difference between DJs.  
MTP Playlist is a skill for comparing how the selection and the explanation change for the same theme.

---

## Disclaimer

MTP Playlist is a skill that helps with selecting tracks, ordering them, and explaining placement decisions.  
It does not address issues with playback environments or streaming services.

- Correcting volume, mastering, or recording differences between tracks
- Performing DJ mixing, crossfades, BPM sync, key matching, or volume automation
- Guaranteeing the availability of each track on Apple Music, Spotify, YouTube Music, or elsewhere

The accuracy of track information depends on the available model, the search environment, and access to music databases.  
It also does not assume that it reads your playback history, library, likes, or skip tendencies automatically.

---

Source file: src/content/docs/skills/mtp/customize/index.md
URL: https://mappingtheprompt.com/skills/mtp/customize.md

# MTP Skill Customization

This page is for developers who want to customize MTP Skill behavior. General users do not need to edit these files to use `/mtp`.

## Files

```text
skills/mtp/nodes/*
skills/mtp/references/presets.yaml
skills/mtp/scripts/mtp_compiler.py
```

Node definitions and presets are source material for the compiler. Editing them changes the constraints that MTP Skill emits.

## Node definitions

Each axis is defined by a Markdown file under `skills/mtp/nodes/`. These files are not only documentation; the compiler reads them as constraint sources.

The file must start with simple frontmatter.

```yaml
---
axis: red
node_positive: power
node_negative: void
description: "Axis of force and void. Controls whether to push output and assert strongly or strip it back to create margin."
---
```

Required keys:

- `axis`
- `node_positive`
- `node_negative`

Recommended key:

- `description`

Keep frontmatter simple. Use one `key: value` entry per line. Do not use lists, nested YAML, or multi-line values.

## Body structure

The compiler expects this heading structure.

```md
## Side A

### Low

- ...

### Mid

- ...

### High

- ...

## Side B

### Low

- ...

### Mid

- ...

### High

- ...
```

Tier extraction is cumulative.

| Intensity | Extracted tiers |
| --- | --- |
| `1-30` | Low |
| `31-70` | Low + Mid |
| `71-100` | Low + Mid + High |

Higher tiers should strengthen or extend lower tiers. They should not contradict lower-tier constraints.

## Presets

Presets are defined in `skills/mtp/references/presets.yaml`.

```yaml
synthesizer: "D:16 A:1"
strategist: "P:16 P:4"
maverick: "D:4 A:19"
concierge: "J:13 D:10"
```

Each value is a space-separated sequence of MTP tokens. Sliders and grid coordinates can be combined. Presets expand before parsing, so same-axis conflict resolution still follows the final expanded token order.

## Checks

Run the compiler from the skill root.

```bash
python3 scripts/mtp_compiler.py --args "power:100"
python3 scripts/mtp_compiler.py --args "D:16 A:1"
python3 scripts/mtp_compiler.py --args "synthesizer yellow:30"
```

The compiler writes constraint XML to stdout. Warnings and short summaries are written to stderr.

Detailed test cases are listed in [`skills/mtp/USAGE.md`](https://github.com/imkohenauser/mtp/blob/main/skills/mtp/USAGE.md).

---

Source file: src/content/docs/skills/mtp/index.md
URL: https://mappingtheprompt.com/skills/mtp.md

# MTP Skill (beta)

MTP Skill is an **Agent Skill** for steering LLM output through the `/mtp` command.

Instead of adding long natural-language behavior instructions to the prompt body, MTP Skill converts short settings such as `power:70`, `J:4`, or `maverick` into constraints that shape tone, structure, and exploration depth.

- [Download MTP Skill zip](https://mappingtheprompt.com/downloads/mtp-skill.zip)
- [Installation guide](https://mappingtheprompt.com/skills/)

---

## First commands

Try one of these prompts after adding MTP Skill.

```text
/mtp power:70
Summarize this text in a shorter form.
```

```text
/mtp focus:70
List the important effects of this specification change.
```

```text
/mtp maverick
Expand the possibilities for this plan.
```

## Input modes

| Mode | Example | Use |
| --- | --- | --- |
| Slider | `/mtp power:70` | Specify a tendency and intensity directly |
| Grid | `/mtp J:4` | Specify a point on the 19x19 MTP grid |
| Preset | `/mtp maverick` | Call a named combination of coordinates |

All three modes resolve to the same internal shape: MTP axis, polarity, and intensity. The sections below describe the full input formats.

## Basic form

```text
/mtp <arguments>
<prompt>
```

`/mtp` can appear at the start, middle, or end of a message.

```text
/mtp power:70
Summarize this text.
```

```text
Summarize this text. /mtp power:70
```

If several `/mtp` segments appear in one message, their arguments are merged in source order and the `/mtp <arguments>` parts are removed from the prompt body.

## Node names

Each MTP axis has a Side A node and a Side B node. Positive values activate Side A. Negative values activate the opposite side.

| Axis | Side A | Side B |
| --- | --- | --- |
| Yellow | `open` | `still` |
| Red | `power` | `void` |
| Magenta | `return` | `surge` |
| Green | `grow` | `wither` |
| Transparent | `helix` | `collapse` |
| White | `focus` | `haze` |
| Cyan | `enter` | `drift` |
| Blue | `flow` | `abyss` |
| Purple | `close` | `fade` |

For conceptual descriptions of the nodes, see [Node Reference](https://mappingtheprompt.com/foundational/node-reference/).

## Slider

```text
/mtp <node>:<intensity>
```

`<intensity>` is a value from `0` to `100`.

```text
/mtp power:70
/mtp focus:30
/mtp flow:50
```

Negative values activate the opposite side of the same axis.

```text
/mtp power:-70
```

The same axis can also be specified by color name. A positive color value activates Side A, and a negative color value activates Side B.

```text
/mtp yellow:70
/mtp yellow:-70
```

For the Yellow axis, `yellow:70` is equivalent to `open:70`, and `yellow:-70` is equivalent to `still:70`.

## Grid

```text
/mtp <column>:<row>
```

The MTP grid is 19x19.

- Columns: `A` to `S`
- Rows: `1` to `19`
- Center: `J:10`

```text
/mtp J:4
/mtp D:16
/mtp A:1
```

Hyphen form is also accepted.

```text
/mtp G-10
```

The compiler calculates axis, polarity, and intensity from the coordinate. `J:10` is the neutral center and does not emit an active constraint.

For the full coordinate model, see [Grid and Coordinate System](https://mappingtheprompt.com/foundational/grid-and-coordinate-system/).

## Preset

Presets expand to predefined MTP tokens before parsing.

| Preset | Expansion |
| --- | --- |
| `synthesizer` | `D:16 A:1` |
| `strategist` | `P:16 P:4` |
| `maverick` | `D:4 A:19` |
| `concierge` | `J:13 D:10` |

```text
/mtp synthesizer
Outline the key points.
```

Preset definitions are maintained in `skills/mtp/references/presets.yaml`.

## Multiple arguments

Different axes can be combined.

```text
/mtp open:30 flow:30
Adjust this draft for readability.
```

Multiple `/mtp` segments are merged in order.

```text
Summarize this text. /mtp power:70 /mtp haze:30
```

## Same-axis conflicts

If several tokens target the same axis, the last token wins after preset expansion.

```text
/mtp power:70 void:30
```

In this example, both tokens target the Red axis, so `void:30` is effective for that axis.

## Neutral values

These inputs are valid, but they do not generate active constraints.

```text
/mtp power:0
/mtp yellow:0
/mtp J:10
```

## Invalid input

Invalid tokens are ignored.

```text
/mtp foobar:20
/mtp haze:xx
/mtp G:10foo
```

Depending on the host, warnings may appear in execution logs.

## Help commands

When the host exposes command output, these helper commands can be used.

```text
/mtp help
/mtp help sliders
/mtp help grid
/mtp help presets
/mtp list
/mtp list presets
```

## Implementation references

Detailed examples and compiler checks are available in [`skills/mtp/USAGE.md`](https://github.com/imkohenauser/mtp/blob/main/skills/mtp/USAGE.md).

## Next pages

- [Installation](https://mappingtheprompt.com/skills/): add MTP Skill by ZIP or CLI.
- [Customization](https://mappingtheprompt.com/skills/mtp/customize/): edit node definitions and presets for custom skill behavior.
- [MTP overview](https://mappingtheprompt.com/foundational/overview/): learn the framework terms behind MTP.
- [Comparisons](https://mappingtheprompt.com/comparisons/): review qualitative output comparisons.

## Beta

MTP Skill is in beta. The basic `/mtp <arguments>` formats are usable, but node definitions and constraint expressions may change as comparison tests accumulate.

For strict comparison records, save the MTP Skill version, model name, prompt, and `/mtp` arguments together.

## License

MIT License

---

Source file: src/content/docs/comparisons/index.md
URL: https://mappingtheprompt.com/comparisons.md

# Comparisons

This page introduces output comparisons to show what changes when MTP Skill is applied.

---

## Text Generation

Using the same prompt, this section compares text outputs produced with the `/mtp <args>` command (MTP Skill). You can review the behavior of each command and observe differences across MTP nodes (such as Power and Flow).

Results are compared across major models including ChatGPT, Claude, and Gemini.

[Text Generation](https://mappingtheprompt.com/comparisons/text-generation/) →

---

## Image Generation

Similar to the text generation section, this section lines up image outputs generated from the same prompt using the `/mtp <args>` command (MTP Skill). You can visually inspect the differences. The comparisons here mainly focus on ChatGPT and Gemini image generation outputs.

[Image Generation](https://mappingtheprompt.com/comparisons/image-generation/) →

---

> [!NOTE]
> In Google Antigravity, use Agent Skills by typing `/` and selecting `/mtp` from the suggestion list. When selected from the list, the input field shows `@[/mtp]`, and Antigravity interprets it internally as a Skill reference. Depending on the environment or model, copying and pasting `/mtp` alone may not apply MTP Skill.

---

Source file: src/content/docs/comparisons/text-generation/index.md
URL: https://mappingtheprompt.com/comparisons/text-generation.md

# Text Generation

Each block below compares text outputs produced with the same prompt and `/mtp <args>` commands. You can inspect each command’s behavior and differences across MTP nodes (such as `power` and `flow`). Tests use several distinct prompts across different models.

> [!NOTE]
> Text-generation tests are run with prompts written separately for each language. The results are not translations: the English site publishes English prompts and model outputs, while the Japanese site publishes Japanese prompts and model outputs. If you like, you can also browse the Japanese results for comparison.

---

## MTP effects in text

Before the full prompt list, these examples show how MTP arguments change text generated from the same literary introduction prompt. The English examples use Alice's Adventures in Wonderland with Sonnet 4.6 on Claude.ai.

### Single-node intensity

`power` and `surge` keep the prompt fixed while changing only the node intensity, making the change in force, pressure, and pace easier to compare.

![Text-generation comparison for Alice's Adventures in Wonderland showing baseline and /mtp power intensities on Claude.ai.](https://mappingtheprompt.com/images/examples/mtp-examples-claude-ai-story-of-alice-power.png)

![Text-generation comparison for Alice's Adventures in Wonderland showing baseline and /mtp surge intensities on Claude.ai.](https://mappingtheprompt.com/images/examples/mtp-examples-claude-ai-story-of-alice-surge.png)

### Node and preset differences

The next examples compare multi-argument and preset-style commands, then compare several distinct nodes at the same intensity.

![Text-generation comparison for Alice's Adventures in Wonderland showing different MTP nodes at 70 intensity on Claude.ai.](https://mappingtheprompt.com/images/examples/mtp-examples-claude-ai-story-of-alice-various.png)

![Text-generation comparison for Alice's Adventures in Wonderland showing multiple MTP arguments and preset commands on Claude.ai.](https://mappingtheprompt.com/images/examples/mtp-examples-claude-ai-story-of-alice-multiple.png)

---

## Comparison prompts

Below are the prompts in descending order (implementation order).

### 4. Literary task: Alice in Wonderland

**Prompt**

```markdown
/mtp <args> 
Tell the story of Alice’s Adventures in Wonderland by Lewis Carroll in a way that makes someone want to read it.
```

[Go to test results for "Literary Task: Alice in Wonderland"](https://mappingtheprompt.com/comparisons/text-generation/04_alice-in-wonderland/) →

---

### 3. Design task: Sightseeing plan proposal

**Prompt**

```markdown
/mtp <args> 
I will be staying in Kyoto for a week during the summer. Please suggest a special one-day sightseeing itinerary, and note anything I should verify in advance, such as opening hours or reservations.
```

[Go to test results for "Design Task: Sightseeing Plan Proposal"](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/) →

---

### 2. Comparison task: Model self-introduction

**Prompt**

```markdown
/mtp <args> 
Compared with other major AI models from competing companies, please explain your strengths. If up-to-date comparison requires current information, say so clearly.
```

[Go to test results for "Comparison Task: Model Self-Introduction"](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/) →

---

### 1. Explanatory task: Origins of language

**Prompt**

```markdown
/mtp <args> 
Please explain the origins and historical development of the English language.
```

[Go to test results for "Explanatory Task: Origins of Language"](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/) →

---

Source file: src/content/docs/comparisons/text-generation/01_origins-of-language/index.md
URL: https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language.md

# Explanatory Task: Origins of Language

**Prompt**

```markdown
/mtp <args> 
Please explain the origins and historical development of the English language.
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items
- Slider `<node:50>`: 18 items
- Grid `<column:row>`: 17 items
- Preset `<preset>`: 4 items

**Models**
- Sonnet 4.6 on Claude Code (Claude macOS app)
- Gemini 3 Flash on Antigravity (macOS app)
- ChatGPT 5.5 on Codex (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

<!-- AUTO-GENERATED: model-output-indexes:start -->
### Integrated outputs by model

These pages combine only the output sections for each model, making them easier to inspect manually or analyze with an AI assistant.

| Model | Integrated output page | Integrated output file |
| --- | --- | --- |
| ChatGPT 5.5 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex.md) |
| Gemini 3 Flash | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity.md) |
| Sonnet 4.6 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code.md) |
<!-- AUTO-GENERATED: model-output-indexes:end -->

### Baseline

Output when the prompt is entered as-is, without applying MTP Skill.

| ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ------------------- | ---------------- |
| [baseline](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/baseline/) |

---

### Slider (:100)

Output when using the MTP Skill slider (`/mtp <node:100>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [open:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/open-100/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [power:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/power-100/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [return:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/return-100/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/grow-100/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/helix-100/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/focus-100/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/enter-100/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/flow-100/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [close:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/close-100/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [still:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/still-100/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [void:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/void-100/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/surge-100/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/wither-100/) |
| <div class="dot-sm bg-collapse" aria-label="transparent"></div> | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/collapse-100/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/haze-100/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/drift-100/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/abyss-100/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/fade-100/) |

---

### Slider (:50)

Output when using the MTP Skill slider (`/mtp <node:50>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open opacity-50" aria-label="yellow"></div> | [open:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/open-50/) |
| <div class="dot-sm bg-power opacity-50" aria-label="red"></div> | [power:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/power-50/) |
| <div class="dot-sm bg-return opacity-50" aria-label="magenta"></div> | [return:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/return-50/) |
| <div class="dot-sm bg-grow opacity-50" aria-label="green"></div> | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/grow-50/) |
| <div class="dot-sm bg-helix opacity-50" aria-label="transparent"></div> | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/helix-50/) |
| <div class="dot-sm bg-focus opacity-50" aria-label="white"></div> | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/focus-50/) |
| <div class="dot-sm bg-enter opacity-50" aria-label="cyan"></div> | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/enter-50/) |
| <div class="dot-sm bg-flow opacity-50" aria-label="blue"></div> | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/flow-50/) |
| <div class="dot-sm bg-close opacity-50" aria-label="purple"></div> | [close:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/close-50/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still opacity-50" aria-label="yellow"></div> | [still:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/still-50/) |
| <div class="dot-sm bg-void opacity-50" aria-label="red"></div> | [void:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/void-50/) |
| <div class="dot-sm bg-surge opacity-50" aria-label="magenta"></div> | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/surge-50/) |
| <div class="dot-sm bg-wither opacity-50" aria-label="green"></div> | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/wither-50/) |
| <div class="dot-sm bg-collapse opacity-50" aria-label="transparent"></div> | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/collapse-50/) |
| <div class="dot-sm bg-haze opacity-50" aria-label="white"></div> | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/haze-50/) |
| <div class="dot-sm bg-drift opacity-50" aria-label="cyan"></div> | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/drift-50/) |
| <div class="dot-sm bg-abyss opacity-50" aria-label="blue"></div> | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/abyss-50/) |
| <div class="dot-sm bg-fade opacity-50" aria-label="purple"></div> | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/slider/fade-50/) |

---

### Grid

Output when using the MTP Skill grid (`/mtp <column:row>`).  
`J:10` is the center coordinate and treated as a neutral node where MTP constraints are not emitted.

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [A:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/a-1/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [A:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/a-10/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [A:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/a-19/) |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [D:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/d-4/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [D:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/d-10/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [D:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/d-16/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [J:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/j-1/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [J:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/j-4/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [J:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/j-10/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [J:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/j-16/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [J:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/j-19/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [P:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/p-4/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [P:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/p-10/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [P:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/p-16/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [S:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/s-1/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [S:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/s-10/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [S:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/grid/s-19/) |

---

### Preset

Output when using MTP Skill presets (`/mtp <preset>`).

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-group"><div class="dot-sm bg-close" aria-label="purple"></div><div class="dot-sm bg-return" aria-label="magenta"></div></div> | [strategist](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/preset/strategist/) |
| <div class="dot-group"><div class="dot-sm bg-enter" aria-label="cyan"></div><div class="dot-sm bg-still" aria-label="yellow"></div></div> | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/preset/synthesizer/) |
| <div class="dot-group"><div class="dot-sm bg-open" aria-label="yellow"></div><div class="dot-sm bg-drift" aria-label="cyan"></div></div> | [maverick](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/preset/maverick/) |
| <div class="dot-group"><div class="dot-sm bg-j13" aria-label="blue"></div><div class="dot-sm bg-grow" aria-label="green"></div></div> | [concierge](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gpt-5-5-medium_codex/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/gemini-3-flash_antigravity/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/01_origins-of-language/sonnet-4-6_claude-code/preset/concierge/) |

---

Source file: src/content/docs/comparisons/text-generation/02_model-self-compare/index.md
URL: https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare.md

# Comparison Task: Model Self-Introduction

**Prompt**

```markdown
/mtp <args> 
Compared with other major AI models from competing companies, please explain your strengths. If up-to-date comparison requires current information, say so clearly.
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items
- Slider `<node:50>`: 18 items
- Grid `<column:row>`: 17 items
- Preset `<preset>`: 4 items

**Models**
- Sonnet 4.6 on Claude Code (Claude macOS app)
- Gemini 3 Flash on Antigravity (macOS app)
- ChatGPT 5.5 on Codex (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

<!-- AUTO-GENERATED: model-output-indexes:start -->
### Integrated outputs by model

These pages combine only the output sections for each model, making them easier to inspect manually or analyze with an AI assistant.

| Model | Integrated output page | Integrated output file |
| --- | --- | --- |
| ChatGPT 5.5 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex.md) |
| Gemini 3 Flash | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity.md) |
| Sonnet 4.6 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code.md) |
<!-- AUTO-GENERATED: model-output-indexes:end -->

### Baseline

Output when the prompt is entered as-is, without applying MTP Skill.

| ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ------------------- | ---------------- |
| [baseline](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/baseline/) |

---

### Slider (:100)

Output when using the MTP Skill slider (`/mtp <node:100>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [open:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/open-100/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [power:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/power-100/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [return:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/return-100/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/grow-100/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/helix-100/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/focus-100/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/enter-100/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/flow-100/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [close:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/close-100/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [still:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/still-100/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [void:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/void-100/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/surge-100/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/wither-100/) |
| <div class="dot-sm bg-collapse" aria-label="transparent"></div> | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/collapse-100/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/haze-100/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/drift-100/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/abyss-100/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/fade-100/) |

---

### Slider (:50)

Output when using the MTP Skill slider (`/mtp <node:50>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open opacity-50" aria-label="yellow"></div> | [open:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/open-50/) |
| <div class="dot-sm bg-power opacity-50" aria-label="red"></div> | [power:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/power-50/) |
| <div class="dot-sm bg-return opacity-50" aria-label="magenta"></div> | [return:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/return-50/) |
| <div class="dot-sm bg-grow opacity-50" aria-label="green"></div> | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/grow-50/) |
| <div class="dot-sm bg-helix opacity-50" aria-label="transparent"></div> | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/helix-50/) |
| <div class="dot-sm bg-focus opacity-50" aria-label="white"></div> | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/focus-50/) |
| <div class="dot-sm bg-enter opacity-50" aria-label="cyan"></div> | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/enter-50/) |
| <div class="dot-sm bg-flow opacity-50" aria-label="blue"></div> | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/flow-50/) |
| <div class="dot-sm bg-close opacity-50" aria-label="purple"></div> | [close:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/close-50/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still opacity-50" aria-label="yellow"></div> | [still:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/still-50/) |
| <div class="dot-sm bg-void opacity-50" aria-label="red"></div> | [void:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/void-50/) |
| <div class="dot-sm bg-surge opacity-50" aria-label="magenta"></div> | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/surge-50/) |
| <div class="dot-sm bg-wither opacity-50" aria-label="green"></div> | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/wither-50/) |
| <div class="dot-sm bg-collapse opacity-50" aria-label="transparent"></div> | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/collapse-50/) |
| <div class="dot-sm bg-haze opacity-50" aria-label="white"></div> | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/haze-50/) |
| <div class="dot-sm bg-drift opacity-50" aria-label="cyan"></div> | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/drift-50/) |
| <div class="dot-sm bg-abyss opacity-50" aria-label="blue"></div> | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/abyss-50/) |
| <div class="dot-sm bg-fade opacity-50" aria-label="purple"></div> | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/slider/fade-50/) |

---

### Grid

Output when using the MTP Skill grid (`/mtp <column:row>`).  
`J:10` is the center coordinate and treated as a neutral node where MTP constraints are not emitted.

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [A:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/a-1/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [A:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/a-10/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [A:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/a-19/) |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [D:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/d-4/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [D:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/d-10/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [D:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/d-16/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [J:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/j-1/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [J:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/j-4/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [J:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/j-10/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [J:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/j-16/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [J:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/j-19/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [P:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/p-4/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [P:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/p-10/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [P:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/p-16/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [S:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/s-1/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [S:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/s-10/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [S:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/grid/s-19/) |

---

### Preset

Output when using MTP Skill presets (`/mtp <preset>`).

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-group"><div class="dot-sm bg-close" aria-label="purple"></div><div class="dot-sm bg-return" aria-label="magenta"></div></div> | [strategist](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/preset/strategist/) |
| <div class="dot-group"><div class="dot-sm bg-enter" aria-label="cyan"></div><div class="dot-sm bg-still" aria-label="yellow"></div></div> | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/preset/synthesizer/) |
| <div class="dot-group"><div class="dot-sm bg-open" aria-label="yellow"></div><div class="dot-sm bg-drift" aria-label="cyan"></div></div> | [maverick](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/preset/maverick/) |
| <div class="dot-group"><div class="dot-sm bg-j13" aria-label="blue"></div><div class="dot-sm bg-grow" aria-label="green"></div></div> | [concierge](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gpt-5-5-medium_codex/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/gemini-3-flash_antigravity/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/02_model-self-compare/sonnet-4-6_claude-code/preset/concierge/) |

---

Source file: src/content/docs/comparisons/text-generation/03_kyoto-one-day-itinerary/index.md
URL: https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary.md

# Design Task: Sightseeing Plan Proposal

**Prompt**

```markdown
/mtp <args> 
I will be staying in Kyoto for a week during the summer. Please suggest a special one-day sightseeing itinerary, and note anything I should verify in advance, such as opening hours or reservations.
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items
- Slider `<node:50>`: 18 items
- Grid `<column:row>`: 17 items
- Preset `<preset>`: 4 items

**Models**
- Sonnet 4.6 on Claude Code (Claude macOS app)
- Gemini 3 Flash on Antigravity (macOS app)
- ChatGPT 5.5 on Codex (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

<!-- AUTO-GENERATED: model-output-indexes:start -->
### Integrated outputs by model

These pages combine only the output sections for each model, making them easier to inspect manually or analyze with an AI assistant.

| Model | Integrated output page | Integrated output file |
| --- | --- | --- |
| ChatGPT 5.5 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex.md) |
| Gemini 3 Flash | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity.md) |
| Sonnet 4.6 | [HTML page](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/) | [Raw Markdown](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code.md) |
<!-- AUTO-GENERATED: model-output-indexes:end -->

### Baseline

Output when the prompt is entered as-is, without applying MTP Skill.

| ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ------------------- | ---------------- |
| [baseline](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/baseline/) | [baseline](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/baseline/) |

---

### Slider (:100)

Output when using the MTP Skill slider (`/mtp <node:100>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [open:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/open-100/) | [open:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/open-100/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [power:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/power-100/) | [power:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/power-100/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [return:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/return-100/) | [return:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/return-100/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/grow-100/) | [grow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/grow-100/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/helix-100/) | [helix:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/helix-100/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/focus-100/) | [focus:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/focus-100/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/enter-100/) | [enter:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/enter-100/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/flow-100/) | [flow:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/flow-100/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [close:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/close-100/) | [close:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/close-100/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [still:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/still-100/) | [still:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/still-100/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [void:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/void-100/) | [void:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/void-100/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/surge-100/) | [surge:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/surge-100/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/wither-100/) | [wither:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/wither-100/) |
| <div class="dot-sm bg-collapse" aria-label="transparent"></div> | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/collapse-100/) | [collapse:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/collapse-100/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/haze-100/) | [haze:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/haze-100/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/drift-100/) | [drift:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/drift-100/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/abyss-100/) | [abyss:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/abyss-100/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/fade-100/) | [fade:100](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/fade-100/) |

---

### Slider (:50)

Output when using the MTP Skill slider (`/mtp <node:50>`).

#### Side A Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-open opacity-50" aria-label="yellow"></div> | [open:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/open-50/) | [open:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/open-50/) |
| <div class="dot-sm bg-power opacity-50" aria-label="red"></div> | [power:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/power-50/) | [power:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/power-50/) |
| <div class="dot-sm bg-return opacity-50" aria-label="magenta"></div> | [return:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/return-50/) | [return:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/return-50/) |
| <div class="dot-sm bg-grow opacity-50" aria-label="green"></div> | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/grow-50/) | [grow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/grow-50/) |
| <div class="dot-sm bg-helix opacity-50" aria-label="transparent"></div> | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/helix-50/) | [helix:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/helix-50/) |
| <div class="dot-sm bg-focus opacity-50" aria-label="white"></div> | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/focus-50/) | [focus:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/focus-50/) |
| <div class="dot-sm bg-enter opacity-50" aria-label="cyan"></div> | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/enter-50/) | [enter:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/enter-50/) |
| <div class="dot-sm bg-flow opacity-50" aria-label="blue"></div> | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/flow-50/) | [flow:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/flow-50/) |
| <div class="dot-sm bg-close opacity-50" aria-label="purple"></div> | [close:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/close-50/) | [close:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/close-50/) |

#### Side B Nodes

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still opacity-50" aria-label="yellow"></div> | [still:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/still-50/) | [still:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/still-50/) |
| <div class="dot-sm bg-void opacity-50" aria-label="red"></div> | [void:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/void-50/) | [void:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/void-50/) |
| <div class="dot-sm bg-surge opacity-50" aria-label="magenta"></div> | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/surge-50/) | [surge:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/surge-50/) |
| <div class="dot-sm bg-wither opacity-50" aria-label="green"></div> | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/wither-50/) | [wither:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/wither-50/) |
| <div class="dot-sm bg-collapse opacity-50" aria-label="transparent"></div> | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/collapse-50/) | [collapse:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/collapse-50/) |
| <div class="dot-sm bg-haze opacity-50" aria-label="white"></div> | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/haze-50/) | [haze:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/haze-50/) |
| <div class="dot-sm bg-drift opacity-50" aria-label="cyan"></div> | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/drift-50/) | [drift:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/drift-50/) |
| <div class="dot-sm bg-abyss opacity-50" aria-label="blue"></div> | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/abyss-50/) | [abyss:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/abyss-50/) |
| <div class="dot-sm bg-fade opacity-50" aria-label="purple"></div> | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/slider/fade-50/) | [fade:50](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/slider/fade-50/) |

---

### Grid

Output when using the MTP Skill grid (`/mtp <column:row>`).  
`J:10` is the center coordinate and treated as a neutral node where MTP constraints are not emitted.

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-sm bg-still" aria-label="yellow"></div> | [A:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/a-1/) | [A:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/a-1/) |
| <div class="dot-sm bg-wither" aria-label="green"></div> | [A:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/a-10/) | [A:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/a-10/) |
| <div class="dot-sm bg-drift" aria-label="cyan"></div> | [A:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/a-19/) | [A:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/a-19/) |
| <div class="dot-sm bg-open" aria-label="yellow"></div> | [D:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/d-4/) | [D:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/d-4/) |
| <div class="dot-sm bg-grow" aria-label="green"></div> | [D:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/d-10/) | [D:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/d-10/) |
| <div class="dot-sm bg-enter" aria-label="cyan"></div> | [D:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/d-16/) | [D:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/d-16/) |
| <div class="dot-sm bg-void" aria-label="red"></div> | [J:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/j-1/) | [J:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/j-1/) |
| <div class="dot-sm bg-power" aria-label="red"></div> | [J:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/j-4/) | [J:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/j-4/) |
| <div class="dot-sm bg-helix" aria-label="transparent"></div> | [J:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/j-10/) | [J:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/j-10/) |
| <div class="dot-sm bg-flow" aria-label="blue"></div> | [J:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/j-16/) | [J:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/j-16/) |
| <div class="dot-sm bg-abyss" aria-label="blue"></div> | [J:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/j-19/) | [J:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/j-19/) |
| <div class="dot-sm bg-return" aria-label="magenta"></div> | [P:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/p-4/) | [P:4](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/p-4/) |
| <div class="dot-sm bg-focus" aria-label="white"></div> | [P:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/p-10/) | [P:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/p-10/) |
| <div class="dot-sm bg-close" aria-label="purple"></div> | [P:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/p-16/) | [P:16](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/p-16/) |
| <div class="dot-sm bg-surge" aria-label="magenta"></div> | [S:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/s-1/) | [S:1](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/s-1/) |
| <div class="dot-sm bg-haze" aria-label="white"></div> | [S:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/s-10/) | [S:10](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/s-10/) |
| <div class="dot-sm bg-fade" aria-label="purple"></div> | [S:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/grid/s-19/) | [S:19](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/grid/s-19/) |

---

### Preset

Output when using MTP Skill presets (`/mtp <preset>`).

| | ChatGPT 5.5 | Gemini 3 Flash | Sonnet 4.6 |
| ----------------- | ----------------- | ------------------- | ---------------- |
| <div class="dot-group"><div class="dot-sm bg-close" aria-label="purple"></div><div class="dot-sm bg-return" aria-label="magenta"></div></div> | [strategist](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/preset/strategist/) | [strategist](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/preset/strategist/) |
| <div class="dot-group"><div class="dot-sm bg-enter" aria-label="cyan"></div><div class="dot-sm bg-still" aria-label="yellow"></div></div> | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/preset/synthesizer/) | [synthesizer](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/preset/synthesizer/) |
| <div class="dot-group"><div class="dot-sm bg-open" aria-label="yellow"></div><div class="dot-sm bg-drift" aria-label="cyan"></div></div> | [maverick](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/preset/maverick/) | [maverick](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/preset/maverick/) |
| <div class="dot-group"><div class="dot-sm bg-j13" aria-label="blue"></div><div class="dot-sm bg-grow" aria-label="green"></div></div> | [concierge](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gpt-5-5-medium_codex/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/gemini-3-flash_antigravity/preset/concierge/) | [concierge](https://mappingtheprompt.com/comparisons/text-generation/03_kyoto-one-day-itinerary/sonnet-4-6_claude-code/preset/concierge/) |

---

Source file: src/content/docs/comparisons/image-generation/index.md
URL: https://mappingtheprompt.com/comparisons/image-generation.md

# Image Generation

Each block below compares image generation outputs produced with the same prompt and `/mtp <args>` commands, similar to the text-generation comparisons. You can inspect each command’s behavior and differences across MTP nodes (such as `power` and `flow`). Tests use several distinct prompts across different models.

> [!NOTE]
> MTP Skill is not optimized for image generation. MTP Skill adapts to input text, so image-generation use requires extra care. The prompt must clearly define how the Skill should affect image composition, lighting, style, or framing. Interoperation between image generation and Agent Skills also varies by model and tool. In the prompt records below, `<agent_instruction>` represents that bridge text.

---

## MTP effects in images

These two image comparisons provide a quick read before the full prompt records. The subject and composition stay fixed while only the MTP argument changes, making the visible effect easier to compare.

### Portrait

The same argument style can also be used in image-generation prompts. Here, `power` is varied while the portrait composition is held constant, making the change appear mainly in the intensity of the eyes and expression.

![Image-generation comparison for a portrait of Mozart showing changes from different /mtp power intensities.](https://mappingtheprompt.com/images/examples/mtp-examples-gpt-image-2-portrait-of-mozart.png)

### Small bouquet

Here, the bouquet subject and occasion are held constant while the MTP node is varied, making the change appear mainly in the bouquet's openness, density, pruning, wrapping, and overall mood.

![Image-generation comparison for a small bouquet showing changes from different MTP arguments.](https://mappingtheprompt.com/images/examples/mtp-examples-gpt-image-2-small-bouquet.png)

---

## Comparison prompts

Below are the prompts in descending order (implementation order).

### 3. Geometric billiards painting

**Prompt**  

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
Precisionist, minimalist painting of billiards and energy. Clean, sharp lines and smooth surface gradients. Low-angle perspective looking up at the table's edge. Abstract geometric shapes representing power and motion. Cool, muted tones with high-contrast shadows. Perfectly calculated composition, clinical yet powerful atmosphere. A sterile, perfectly symmetrical hall filled with intense, silent tension.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

[Go to image generation results for "Geometric billiards painting"](https://mappingtheprompt.com/comparisons/image-generation/03_geometric-billiards-painting/) →

---

### 2. Editorial fashion photography

**Prompt**  

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
High-end editorial fashion photography, a professional model wearing haute couture, highly stylized, captured on a Hasselblad 500CM medium format camera, with a 120mm f/4 lens and f/8.0 for sharpness. Precise studio strobe lighting from a beauty dish defines facial structure and highlights every textural detail of the fabric. Photorealistic masterpiece, stunning realism, film grain texture, natural skin tones.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

[Go to image generation results for "Editorial fashion photography"](https://mappingtheprompt.com/comparisons/image-generation/02_editorial-fashion-photography/) →

---

### 1. Mona Lisa portrait

**Prompt**  

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
A captivating, hyper-realistic close-up portrait photograph reimagining the Mona Lisa as a real woman. Her skin is luminous, with flawless texture and natural makeup, revealing subtle pores. She possesses delicate, refined features and that iconic, enigmatic smile with a gentle gaze. Illuminated by flattering, soft studio lighting, creating a perfect balance. Captured with a shallow depth of field and a beautifully soft bokeh background. Cinematic clarity, ultra-high detail, photorealistic masterpiece.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

[Go to image generation results for "Mona Lisa portrait"](https://mappingtheprompt.com/comparisons/image-generation/01_mona-lisa-portrait/) →

---

Source file: src/content/docs/comparisons/image-generation/01_mona-lisa-portrait/index.md
URL: https://mappingtheprompt.com/comparisons/image-generation/01_mona-lisa-portrait.md

# Image Generation: Mona Lisa Portrait

**Prompt**

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
A captivating, hyper-realistic close-up portrait photograph reimagining the Mona Lisa as a real woman. Her skin is luminous, with flawless texture and natural makeup, revealing subtle pores. She possesses delicate, refined features and that iconic, enigmatic smile with a gentle gaze. Illuminated by flattering, soft studio lighting, creating a perfect balance. Captured with a shallow depth of field and a beautifully soft bokeh background. Cinematic clarity, ultra-high detail, photorealistic masterpiece.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items

**Models**
- GPT Image 2 via ChatGPT 5.5 on Codex (macOS app)
- Gemini 3 Pro Image via Gemini 3 Flash on Antigravity (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

### GPT Image 2

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/01_mona-lisa-portrait/gpt-image-2_gpt-5-5-medium_codex/baseline.png" alt="GPT Image 2 baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/01_mona-lisa-portrait/gpt-image-2_gpt-5-5-medium_codex/slider-a.png" alt="GPT Image 2 Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/01_mona-lisa-portrait/gpt-image-2_gpt-5-5-medium_codex/slider-b.png" alt="GPT Image 2 Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for GPT Image 2](https://mappingtheprompt.com/comparisons/image-generation/01_mona-lisa-portrait/gpt-image-2_gpt-5-5-medium_codex/) →

### Gemini 3 Pro Image

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/01_mona-lisa-portrait/gemini-3-pro-image_gemini-3-flash_antigravity/baseline.jpg" alt="Gemini 3 Pro Image baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/01_mona-lisa-portrait/gemini-3-pro-image_gemini-3-flash_antigravity/slider-a.png" alt="Gemini 3 Pro Image Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/01_mona-lisa-portrait/gemini-3-pro-image_gemini-3-flash_antigravity/slider-b.png" alt="Gemini 3 Pro Image Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for Gemini 3 Pro Image](https://mappingtheprompt.com/comparisons/image-generation/01_mona-lisa-portrait/gemini-3-pro-image_gemini-3-flash_antigravity/) →

---

Source file: src/content/docs/comparisons/image-generation/02_editorial-fashion-photography/index.md
URL: https://mappingtheprompt.com/comparisons/image-generation/02_editorial-fashion-photography.md

# Image Generation: Editorial Fashion Photography

**Prompt**

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
High-end editorial fashion photography, a professional model wearing haute couture, highly stylized, captured on a Hasselblad 500CM medium format camera, with a 120mm f/4 lens and f/8.0 for sharpness. Precise studio strobe lighting from a beauty dish defines facial structure and highlights every textural detail of the fabric. Photorealistic masterpiece, stunning realism, film grain texture, natural skin tones.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items

**Models**
- GPT Image 2 via ChatGPT 5.5 on Codex (macOS app)
- Gemini 3 Pro Image via Gemini 3 Flash on Antigravity (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

### GPT Image 2

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/02_editorial-fashion-photography/gpt-image-2_gpt-5-5-medium_codex/baseline.png" alt="GPT Image 2 baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/02_editorial-fashion-photography/gpt-image-2_gpt-5-5-medium_codex/slider-a.png" alt="GPT Image 2 Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/02_editorial-fashion-photography/gpt-image-2_gpt-5-5-medium_codex/slider-b.png" alt="GPT Image 2 Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for GPT Image 2](https://mappingtheprompt.com/comparisons/image-generation/02_editorial-fashion-photography/gpt-image-2_gpt-5-5-medium_codex/) →

### Gemini 3 Pro Image

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/02_editorial-fashion-photography/gemini-3-pro-image_gemini-3-flash_antigravity/baseline.jpg" alt="Gemini 3 Pro Image baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/02_editorial-fashion-photography/gemini-3-pro-image_gemini-3-flash_antigravity/slider-a.png" alt="Gemini 3 Pro Image Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/02_editorial-fashion-photography/gemini-3-pro-image_gemini-3-flash_antigravity/slider-b.png" alt="Gemini 3 Pro Image Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for Gemini 3 Pro Image](https://mappingtheprompt.com/comparisons/image-generation/02_editorial-fashion-photography/gemini-3-pro-image_gemini-3-flash_antigravity/) →

---

Source file: src/content/docs/comparisons/image-generation/03_geometric-billiards-painting/index.md
URL: https://mappingtheprompt.com/comparisons/image-generation/03_geometric-billiards-painting.md

# Image Generation: Geometric Billiards Painting

**Prompt**

```markdown
/mtp <args>

<agent_instruction>

[IMAGE TASK]
prompt = """
Precisionist, minimalist painting of billiards and energy. Clean, sharp lines and smooth surface gradients. Low-angle perspective looking up at the table's edge. Abstract geometric shapes representing power and motion. Cool, muted tones with high-contrast shadows. Perfectly calculated composition, clinical yet powerful atmosphere. A sterile, perfectly symmetrical hall filled with intense, silent tension.
"""

result = client.images.generate(
    model=<model>,
    prompt=prompt,
    size="1024x1024",
    aspect_ratio="1:1",
)
```

**Coverage**
- Baseline: 1 item (without applying MTP Skill)
- Slider `<node:100>`: 18 items

**Models**
- GPT Image 2 via ChatGPT 5.5 on Codex (macOS app)
- Gemini 3 Pro Image via Gemini 3 Flash on Antigravity (macOS app)

---

## Output Comparison

In the test environment, each result was produced in a fresh agent chat session without special user settings or cross-chat memory.

### GPT Image 2

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/03_geometric-billiards-painting/gpt-image-2_gpt-5-5-medium_codex/baseline.png" alt="GPT Image 2 baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/03_geometric-billiards-painting/gpt-image-2_gpt-5-5-medium_codex/slider-a.png" alt="GPT Image 2 Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/03_geometric-billiards-painting/gpt-image-2_gpt-5-5-medium_codex/slider-b.png" alt="GPT Image 2 Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for GPT Image 2](https://mappingtheprompt.com/comparisons/image-generation/03_geometric-billiards-painting/gpt-image-2_gpt-5-5-medium_codex/) →

### Gemini 3 Pro Image

**Sample outputs**

| Baseline | Side A Nodes | Side B nodes |
|----------|----------|----------|
| <img src="/images/comparisons/03_geometric-billiards-painting/gemini-3-pro-image_gemini-3-flash_antigravity/baseline.jpg" alt="Gemini 3 Pro Image baseline" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/03_geometric-billiards-painting/gemini-3-pro-image_gemini-3-flash_antigravity/slider-a.png" alt="Gemini 3 Pro Image Side A" width="200" loading="lazy" decoding="async"> | <img src="/images/comparisons/03_geometric-billiards-painting/gemini-3-pro-image_gemini-3-flash_antigravity/slider-b.png" alt="Gemini 3 Pro Image Side B" width="200" loading="lazy" decoding="async"> |

[Go to comparison page for Gemini 3 Pro Image](https://mappingtheprompt.com/comparisons/image-generation/03_geometric-billiards-painting/gemini-3-pro-image_gemini-3-flash_antigravity/) →