# Dialectic Press

> Transform dialectic outputs into essays — meso-level AI-assisted writing with structural recipes. Use after completing a dialectic when you want to turn the analysis into a publishable essay that exposes both the insight and the method.

- Skill: `kyleamathews/dialectic-press` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add kyleamathews/dialectic-press`
- Raw SKILL.md: https://api.skillmd.com/api/skills/kyleamathews/dialectic-press/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: kyleamathews (https://skillmd.com/u/kyleamathews)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/kyleamathews/dialectic-press

---


# Dialectic Press

Transform dialectic outputs into polished essays — AI-assisted writing at "meso" abstraction with a structural recipe at the end.

## When to Use

- You've completed a dialectic and want to turn it into a shareable essay
- The dialectic produced genuine insight — a reframing, a hidden question, a synthesis
- You want the essay to be **pleasurable, not just summarizable**
- You want to expose the method (the recipe) so readers can apply similar moves

## The Method

1. **Provocation/Hook** — Concrete phenomenon that illustrates something larger
2. **Reframing via analogy** — Map onto a well-understood system from another domain
3. **Conceptual currency** — Coin precise terms that do load-bearing work
4. **Structural unpacking** — Decompose into layers, phases, or tensions
5. **Implications/Stakes** — What follows if the reframing is correct?
6. **Recipe** — Expose the generative moves as imperative verbs

What makes this different:
- **Meso-level density** — more detail than humans hold unassisted, less than AI handles autonomously
- **Structural transparency** — the recipe says "here's how the sausage was made"
- **Human brings the moves** — reframe, unpack, trace propagation, coin term
- **AI executes at density** — fills implications, generates prose holding many threads

---

## Process Overview

This is an **interactive workshop**, not an autonomous pipeline. Each phase ends with a checkpoint where the user validates or redirects.

```
Phase 0: Intent       → User establishes what they want to write about
Phase 1: Inventory    → User confirms: "These are the key elements"
Phase 2: Analogy      → User picks from candidate analogies
Phase 3: Concepts     → User validates which terms do work
Phase 4: Outline      → User approves arc before writing
Phase 5: Write & Validate → Draft produced, cleaned, validated, user reviews
```

**IMPORTANT:** Read each phase file only when you reach that phase. Do not read ahead. This keeps instructions fresh in context and prevents skipping checkpoints.

---

## Starting the Process

1. Ask the user which dialectic they want to draw from
2. Read `phases/0-intent.md`
3. Execute Phase 0 — establish what they actually want to write (may differ from dialectic topic)
4. Only after user confirms intent, read `phases/1-inventory.md`
5. Continue phase by phase

The user can intervene at any checkpoint — redirecting, adding constraints, or returning to an earlier phase.

---

## Phase Files

- `phases/0-intent.md` — Establish what the essay is actually about
- `phases/1-inventory.md` — Survey dialectic outputs, identify key elements
- `phases/2-analogy.md` — Find and propose analogical domains
- `phases/3-concepts.md` — Coin and validate conceptual currency
- `phases/4-outline.md` — Map material onto the essay arc
- `phases/5-write.md` — Write, validate, and finalize

---

## Recipe Variations

| Style | When to Use |
|-------|-------------|
| **Numbered conceptual moves** | Essay follows a clear build-up sequence |
| **Process description** | Work was more exploratory ("scan → reframe → structure → style") |
| **Co-authoring protocol** | Want to expose the human-AI collaboration |
| **Honest mess** | Process was genuinely messy |