Moonborn — Developers

Use cases

Verticals where character consistency, voice fingerprinting, lineage, and ensemble relationships compound into product value. Honest scope — Moonborn is the conversational layer, not the whole product.

What this page is

Moonborn is a character consistency engine. The verticals below are the patterns we see teams build with it. Each guide names the capabilities involved, the minimum integration, and the tier you'd need.

Use cases without a guide yet are listed for visibility — they're scoped in v1 of the product but a dedicated walkthrough lands in a follow-up release.

Pick a vertical

Character-consistent support

Drift detection, voice fingerprinting, and long-term memory — wired into a customer support agent so the brand voice holds across thousands of sessions and QA gets alerted when it slips.

Brand voice variants

Lineage and fork-tree governance for brand teams that need formal vs casual, B2B vs B2C, or EN vs DE variants of the same brand persona — without losing the through-line.

Multi-character scenes

Ensemble relationships between personas — for UX research panels, creative writing dialogue partners, internal simulation tooling, and interactive fiction prototypes. Moonborn provides the character layer; orchestration is yours.

Drop in — OpenAI and MCP

Two integration paths — an OpenAI-compatible endpoint for apps already on the OpenAI SDK, and an MCP server that exposes Moonborn personas as Model Context Protocol resources to IDEs and agent frameworks.

Marketplace publishing

List a persona on the Moonborn marketplace with pre-publish moderation, fork lineage with attribution, ratings, and an optional commerce layer via Stripe Connect. Free workspaces can browse and fork; publishing starts at Pro.

Persona-as-code

Sync personas to a Git repository as YAML+markdown files. Pull, push, or bidirectional, scoped to a base path, with last-write-wins conflict detection via Git blob SHA. Team tier and up.

Quality pipeline

Three runtime quality gates — an LLM-as-judge audit across five dimensions, a 33-test provocation suite, and cosine-distance distinctiveness against a baseline. Wired into generation and edit by default; queryable on demand for QA.

Synthetic user research

Run qualitative research panels at the speed of code. Five-persona panels, structured interviews, longitudinal consistency — without the recruiting overhead.

Game NPC orchestration

Persistent, in-character NPCs across long play sessions. Multi-character scenes with ensemble relationships. Voice drift checks so the wizard doesn't slowly turn into a sarcastic AI.

Multi-locale brand persona

One canonical brand voice, six locale variants. Lineage governance, distinctiveness floors against the canonical, audit-trail in every market.

Enterprise RBAC + SSO

SAML 2.0 single sign-on, SCIM 2.0 provisioning, custom roles, IP allowlists, and data-residency lock — the org-level identity story for regulated buyers.

Audit + compliance

GDPR DSARs, account deletion, retention policies, data residency, hash-chained audit log. The compliance surface Moonborn ships out of the box.

Capability map

Capability	Powers
4-layer schema (Soul/Self/Mask/Surface)	All use cases — the substrate
Voice fingerprint + drift detection	Support, brand voice variants
Lineage + fork tree	Brand variants, marketplace publishing
Ensemble relationships	Multi-character scenes
Long-term memory + retrieval	Support, ongoing conversations
LLM-as-judge audit + provocation tests	Quality pipeline, support QA
OpenAI-compatible endpoint	Drop-in for existing apps
MCP server	IDE / agent framework integrations
Git sync	Persona-as-code
Marketplace commerce (Stripe Connect)	Creator monetization

Every row maps to features scoped in v1; nothing here is on a wish list.

What Moonborn is not

Moonborn powers the conversational layer. The surrounding workflow — ticketing, course curriculum, game engine integration, CRM pipelines, scene state machines — is yours.

Teams in game development, sales coaching, education, IT helpdesk, and voice licensing have all used Moonborn via the Chat API or MCP server. We don't pretend to be a game engine, a learning management system, or a CRM. Pick the layer that fits and build the rest of the product around it.

Pick the right tier

Vertical	Minimum tier
Drop-in OpenAI swap, basic chat	Free
Character-consistent support, brand variants, ensembles	Pro
Persona-as-code, MCP server, custom audit tests	Team
Data residency lock, SAML/SCIM, custom contracts	Enterprise

See the full tier matrix for the per-row breakdown.