SenTeGuard logo
Start a free scan
← Back to Home

Joseki: WrapperHub

Open App →

A registry for versioned, tested, and governed LLM behavior packages: prompts, fine-tunes, adapters, evals, and safety policies packaged as one reusable, auditable unit.

Built by SenTeGuard. Currently in early access: join the waitlist for Summer 2026.

Read our intro post: Meet Joseki →

The Problem

AI teams have outgrown one-off prompts. Prompts live in docs. Fine-tune configs sit in notebooks. Eval suites are buried in repos. Safety rules exist only in vendor dashboards.

Every model switch means rebuilding from scratch. Every handoff means copying prompts into Slack. Every audit means digging through git history. There is no standard way to package, version, or share the full behavior of an LLM deployment.

The Solution

Joseki WrapperHub lets you create Behavior Packages: versioned bundles that capture everything about how an LLM should behave:

  • PromptSpec: System prompts, templates, and examples
  • FineTuneSpec: Training recipes and configurations
  • EvalSpec: Evaluation suites and benchmarks
  • SafetySpec: Safety policies and guardrails
  • LicenseSpec: Usage terms and attribution
  • EvidencePackage: Provenance and audit artifacts

Key Features

  • Liveness Score

    Every package carries a recency-weighted "Works / Broken" score per model, updated in real time by the community. Know immediately which models a prompt is alive on: and which ones broke this morning.

  • Versioned Registry

    Install, inspect, and evaluate behavior packages before using them in production. Package signing, artifact hashes, and license metadata travel with every version for tamper-proof integrity.

  • Reputation System

    Creators who maintain their packages earn reputation that compounds into a Go-style kyu / dan rank: the same way Stack Overflow rewards consistent contributors.

  • Data Packs

    Scoped datasets bundled alongside prompts. Versioned, hashed, and licensed like code so the data your model needs is as portable as the prompt itself.

  • CLI First

    joseki install, joseki inspect, and joseki eval bring the package management experience developers already know to LLM behavior.

  • Governance Built In

    Automated safety checks on every version publish. Evidence packs for audit and provenance. License metadata that travels with every download.

Related work

WrapperHub is related work from SenTeGuard: package intended model behavior so a team can share a version, not a one-off prompt. The customer products are MoyoMap and Moyo Scan.

Get Early Access

Joseki:WrapperHub is launching Summer 2026. Join the waitlist to be among the first to publish and install behavior packages.

Visit Joseki →