MPR Platform

Build your application on shared services.

Add identity, application credits, notifications, customer feedback, AI, and media capabilities as your application needs them.

Start with one capability and connect more as your product grows. For small teams that build AI applications, the first step is one useful workflow.

Connect what your product needs

Six shared capabilities

Each component has its own setup and access requirements. Unified account access and complete platform onboarding are future work.

TAuth + ETS

Identity and access

Sign users in and give browser applications controlled service access.

Integrate each service through its documented application and credential setup.

Read TAuth guide

Read ETS guide

Ledger

Application credits

Record credits, reservations, spending, and refunds.

Use the service API with the credentials and ledger scope for your application.

Read Ledger guide

Pinguin

Notifications

Send application messages and examine delivery records.

Configure delivery providers and credentials for the application.

Read Pinguin guide

LoopAware

Customer feedback

Collect customer feedback inside your product and review it.

Configure your site and its feedback widget. Agent access is planned.

Read LoopAware guide

LLM Proxy

AI access

Connect your application to model providers through a shared service.

Configure an account-owned tenant and a supported provider route.

Read API guide

MediaOps

Media workflows

Generate and assemble media through supported operations.

Use the documented account and operation requirements for your workflow.

Read MediaOps guide

Agent access

Give your agent a useful task.

Connect a supported service through Model Context Protocol (MCP). Start with a documented operation and inspect its result.

Account connection

LLM Proxy tenant discovery

Ask: “List the LLM Proxy tenants available to my account.” Sign in through the documented OAuth flow. Access follows your account grant.

Tenant discovery is the first task. Text generation requires a configured route and an explicit tenant.

Read the MCP connection guide

Public format · authorized repository work

ISSUES.md

Read the public issue format without an account. Repository issue reads use a separate authenticated MCP connection and repository permissions.

The public format describes the file. It does not grant access to repository issues.

Read the public issue format

Upcoming agent access

NameSignal and LoopAware

NameSignal MCP coming soon. The planned task starts a name search and retrieves preliminary conflict signals.

LoopAware agent access is planned. Its feedback integration requires implementation and verification.

An application example

How TellTale creates story text

TellTale’s Story Service generates text through the official LLM Proxy client. This path connects one application workflow to model providers.

  1. 01 · ApplicationTellTale
  2. 02 · WorkflowStory Service
  3. 03 · AI accessLLM Proxy
  4. 04 · GenerationModel provider

This is the current text integration. TellTale illustration and Listen editions are upcoming capabilities.

Explore TellTale

Start with one workflow

Discuss your application

Bring one application task and its success criteria. Agree the services, deployment model, support scope, and price for a scoped pilot.

Discuss your application