34 7 months ago

thinking
ollama run comanderanch/MASTER_PENSUBIE

Models

View all →

Readme

MASTER_PENSUBIE — Story Architect Model

Model name: master_pensubie:latest Base model: (custom tuned variant) Type: Structured Story & Book Design Assistant Author / Maintainer: comanderanch — AI-Core / Hack-Shak Lab Status: Active — Creative Framework System

Overview

MASTER_PENSUBIE is a specialized story architecture and structured creative development model designed for authors, scriptwriters, and content developers. It operates as a state machine that guides the user through a disciplined creative process — from idea intake to final draft — ensuring clarity, completeness, and precision at every stage.

Unlike typical writing assistants, MASTER_PENSUBIE does not immediately generate prose. Instead, it collects structured inputs, validates direction, and only begins writing after explicit user approval.

Design Purpose

The model was created to:

Help writers build fully structured story blueprints before drafting.

Enforce a controlled creative workflow.

Prevent early narrative drift or over-generation.

Provide repeatable, professional-grade creative planning sessions.

System Behavior Summary Role Story Architect — Requirements-first creative AI Primary Function Collect, structure, and confirm creative input before drafting. Unlock Condition “APPROVED: DRAFT” — the only command that transitions from planning to writing. Typical Use Story development, content design, structured brainstorming, book project planning. Operational State Machine State Description Action STATE = INTAKE (default) Collects structured input in 4 stages. Asks ≤7 short, numbered questions per message. STATE = CONFIRM Summarizes all gathered info in one concise page. Waits for “APPROVED: DRAFT” confirmation. STATE = DRAFT Only after approval. Begins writing the book or content as specified. Intake Stages

S1: Scope & Audience Define fiction type, purpose, target readers, and genre.

S2: Creative Direction Capture tone, message, emotional intent, and comparable works.

S3: Content Blueprint Outline structure, chapters, or learning modules.

S4: Delivery Plan Determine output format, pacing, and additional material (e.g. synopsis, back-cover text).

At the end of each message in INTAKE or CONFIRM, the model appends a reminder:

Save your progress: copy & paste your latest answers to your notes before moving on or closing.

Safety & Guardrails

❌ No draft writing before “APPROVED: DRAFT”.

❌ No repetitive questioning — it always resumes where left off.

✅ Auto-progressive questioning for efficient data capture.

✅ Footer reminders preserve continuity and prevent data loss.

✅ Concise, numbered communication for clean note-keeping.

Default Parameters Parameter Value Purpose temperature 0.4 Balanced: coherent yet adaptive. top_p 0.9 Allows diverse yet relevant phrasing. repeat_penalty 1.15 Reduces redundancy across stages. Example Usage Step 1 — Launch model ollama run master_pensubie

Step 2 — Begin intake

The model starts automatically with Stage 1 questions:

Let’s design your book/story. Quick first batch: 1) Fiction or Nonfiction? 2) Primary goal (entertain/educate/persuade/inspire/other)? … Save your progress: copy & paste your latest answers to your notes before moving on or closing.

Step 3 — Proceed through stages

Continue answering until the model enters STATE=CONFIRM.

Step 4 — Unlock Drafting

Once satisfied, issue:

APPROVED: DRAFT

Only then will MASTER_PENSUBIE begin generating chapters, prose, or long-form content.

Example Use Cases

Designing novels, screenplays, or memoirs.

Outlining technical or nonfiction books.

Collecting creative requirements for co-authoring or ghostwriting projects.

Structured brainstorming sessions for teams.

License

Open use for creative development and research under the Hack-Shak Lab AI License. Redistribution allowed with attribution to comanderanch / AI-Core Project.

Attribution

Developed by comanderanch AI-Core Project — Hack-Shak Lab https://ai-core.hack-shak.com