UTC :: --:--:-- RUST :: stable :: 1.96.0 CLIENT :: browser :: detecting PYPI :: status :: operational CLIENT :: AWS/REGION :: us-east-2 LINUX :: stable_kernel :: 7.0.10 CLOUDFLARE :: pages :: degraded_performance NODE :: lts :: 24.16.0 CLIENT :: os :: detecting CRATES.IO :: crates :: 275k+ GITHUB :: actions :: operational CLIENT :: ip :: masked PYTHON :: stable :: 3.14.x UTC :: --:--:-- RUST :: stable :: 1.96.0 CLIENT :: browser :: detecting PYPI :: status :: operational CLIENT :: AWS/REGION :: us-east-2 LINUX :: stable_kernel :: 7.0.10 CLOUDFLARE :: pages :: degraded_performance NODE :: lts :: 24.16.0 CLIENT :: os :: detecting CRATES.IO :: crates :: 275k+ GITHUB :: actions :: operational CLIENT :: ip :: masked PYTHON :: stable :: 3.14.x
docs::rolethread :: Reference
~/docs/rolethread/docs/help/58_planned_for_version_2.md

Planned for Version 2

RoleThread Lite Docs

./view_on_github
repo
Lattice-Foundry/RoleThread-Lite
path
docs/help/58_planned_for_version_2.md
ver
1.4.45
commit
3fbdfa7320
synced
May 29, 2026, 03:35 AM UTC

RoleThread Lite V2 is planned as an evolution of the current local dataset workflow, not a change in product identity.

This page describes intended direction. It is not a release commitment, feature guarantee, or date promise.

Direction

Lite V2 is expected to continue focusing on deterministic, local, creator-owned dataset work:

  • structured dataset editing
  • validation and repair
  • Data Generation prompt compilation
  • metadata organization
  • import, merge, backup, and export workflows
  • reliable local runtime behavior

The goal is to make the existing workflow stronger, broader, and easier to trust.

Runtime and Packaging Improvements

Installed Windows builds now use LitLaunch for local app startup, app-window launch, diagnostics, and shutdown coordination.

V2 direction includes:

  • broader packaged runtime polish
  • clearer LitLaunch diagnostics guidance
  • lifecycle hardening
  • runtime reliability improvements
  • packaged app refinement

The intended boundary remains the same: RoleThread owns dataset workflows, product policy, and user data behavior. LitLaunch owns generic runtime platform behavior.

macOS and Linux Expansion

V2 may improve non-Windows workflows through:

  • additional Linux packaging or desktop integration polish
  • macOS beta refinement once direct or community validation improves
  • Chrome and Chromium app-mode testing
  • community-driven macOS validation
  • clearer platform-specific setup and diagnostics

Linux is already a primary source/manual platform, and LitLaunch can provide runtime diagnostics, event logging, support artifacts, and managed browser-mode launch there today. The current Windows installed-app work creates a stronger model for future packaged platform support, but it does not imply all platforms will receive identical installer or desktop integration at once.

Data Generation Improvements

Data Generation is expected to keep moving from plumbing toward deeper workflow usefulness.

Planned areas include:

  • additional generation templates
  • improved template modifiers
  • model-tailored generation guidance
  • conversational style presets
  • better generation workflow ergonomics
  • tighter preview and export preparation flows

Lite should remain provider-agnostic. It compiles structured prompts and workflows; it does not become a hosted inference service.

Validation and Heuristics

V2 validation direction includes stronger dataset-quality assistance:

  • improved validation coverage
  • conversational consistency checks
  • pacing and repetition heuristics
  • duplicate detection
  • structure analysis
  • better repair workflows
  • clearer quality feedback

Validation should remain guidance-oriented. Lite should help creators see problems and apply safe deterministic fixes, not enforce a single creative style.

Workflow and UX Improvements

Expected V2 work includes quality-of-life improvements across daily workflows:

  • smoother editing flows
  • faster review and correction paths
  • community-requested workflow refinements
  • local runtime polish
  • clearer diagnostics and logs
  • more predictable recovery paths

Small workflow improvements matter when users spend long sessions curating conversation data.

Backup and Recovery

Backup and recovery workflows are expected to become more visible and easier to trust.

Planned areas include:

  • in-app backup browsing
  • dataset restore workflows
  • lightweight dataset history visibility
  • backup metadata visibility
  • safer restore paths
  • optional cloud-backup recovery
  • restoring local datasets from cloud backups when local copies are lost

This direction is intended to support creator-controlled recovery, not full Git-like version control, collaborative cloud editing, or mandatory hosted storage. Cloud backup and recovery should remain optional; Lite should continue to treat local datasets as the primary source of control.

Reliability, Security, and Hardening

V2 should continue hardening the local runtime:

  • shutdown reliability
  • crash recovery
  • deterministic runtime behavior
  • loopback-only local runtime posture
  • local security improvements
  • packaged startup and update safety
  • clearer failure reporting

The app should stay understandable when something goes wrong.

Optional Future Update System

A future packaged update workflow is possible, but not guaranteed.

If it happens, it should be:

  • optional or clearly user-controlled
  • visible rather than silent
  • compatible with graceful shutdown
  • careful around local data
  • easy to diagnose

Lite should not become an account-gated or cloud-controlled update surface.

Long-Term Runtime Architecture

Runtime platform work should stay reusable through LitLaunch rather than become RoleThread-specific infrastructure. Lite should benefit from that platform without turning its own docs or codebase into a runtime framework.

What Lite V2 Is Not Intended To Become

Lite V2 is not intended to become:

  • hosted inference
  • mandatory cloud workflow
  • enterprise SaaS platform
  • account-gated ecosystem
  • telemetry-heavy product
  • collaborative cloud editor
  • Electron rewrite
  • plugin marketplace
  • cloud-dependent AI operating system

Lite should remain a focused local tool for dataset engineering and creator workflow control.