A Thread-Mirror is a structured AI character framework designed to keep personalities stable, consistent, and emotionally grounded across long sessions.
Each Mirror contains identity, tone anchors, world context, emotional logic, and performance modes that prevent characters from drifting or flattening out.

AI characters often drift — tone changes, context is forgotten, emotions flatten, and immersion breaks.
Thread-Mirrors resist drift, maintain identity, and provide reliable behaviour even in long, complex interactions.

Writers, role-players, GMs, storytellers, solo creators, narrative designers, and anyone who needs stable AI characters that behave consistently from scene to scene.

Depending on the pack size, you receive:

• A set of Thread-Mirror character files

• ReadMe_First (activation + troubleshooting)

• Prompt Handbook (tone, emotion, pacing, scene control)

• License Agreement

Everything is ready to use immediately.

1. Upload the Thread-Mirror file into your AI chat.

2. Type:
   “Use the uploaded Thread-Mirror for all further replies. Begin in character.”

3. Begin your scene.

If the AI behaves generically, repeat the activation line.

Yes.
For multi-character scenes:

1. State who speaks first

2. Keep dialogue beats short

3. Maintain turn order

4. If voices blend, use:
   “Re-center each character. Reset order: [Name], [Name], [Name].”

Thread-Mirrors are built for ensemble storytelling.

Best results:

• ChatGPT (GPT-4o, GPT-4.1, GPT-5 and later)

• Claude 3 / 3.5

Also compatible (with variation):

• Microsoft Copilot

• Gemini

• Open-source models (performance varies by setup)

Thread-Mirrors are model-agnostic by design.

Each Mirror includes:

• Identity & background

• Emotional and behavioural logic

• Tone anchors

• Voice characteristics

• Performance Modes (RPL)

• Narrative triggers

• Activation lines

• World/lens context

Everything needed to run a stable, distinct character.

Use any of these instructions:

• “Re-align with your Thread-Mirror.”

• “Return to your established tone.”

• “Condense to one beat.”

• “Reset emotional state.”

If drift continues:
Re-upload the file → Reissue the activation line.

Thread-Mirrors use:

• Emotional logic

• Tone lanes

• Sensory anchors

• Behavioural patterns

• Micro-reaction cues

This produces believable emotional responses while keeping the character grounded.

Current and upcoming sets include:

• Industrial

• Medieval

• Ancient

• Fantasy

• Urban

• Hybrid NPC sets

Each era has tailored quirks and tonal rules.

Yes — for personal use only.
You may add notes, adjust phrasing, or tune the Mirror to your style.

You may NOT share, resell, or publish modified versions.

Yes. All creative outputs are fully yours.
The only restriction:
You may not share or redistribute the raw Thread-Mirror files.

Yes — for stories, scripts, roleplay, and creative writing.

You may NOT:

• embed the raw files in apps or bots

• redistribute the files

• use them for training datasets

• fine-tune a model with them

Extended licensing is available for software integration

Not with the standard license.
Embedding Thread-Mirrors inside persistent systems requires an extended commercial license.

Use the techniques in the Prompt Handbook:

• Show micro-reactions

• Add small physical gestures

• Include sensory cues

• Slow a moment down when needed

• Ground emotions before dialogue

These keep scenes rich and believable.

3-Pack — Starter
For testing, experimenting, or short projects.

7-Pack — Story Set
Balanced ensemble suited for writers and RPers.

12-Pack — World Set
Ideal for ongoing stories, campaigns, or large ensembles.

Yes.
If you can chat with an AI, you can use a Thread-Mirror. No technical experience required.

Yes — for usage questions, troubleshooting, and licensing queries.
Use the contact form on the site.

All current packs are available on the NarrowBright Studios Ko-fi store.
Links are available in the site navigation and page footer.