v1.3.3 — 2026-01-04
- Updated root README to reflect Modelfile-based execution surface
- Clarified
Modelfile-rouheas the sole active Modelfile - Locked archive structure and naming conventions
- No behavioral or architectural changes
v1.3.2 — 2026-01-04
- Finalized and archived
Modelfile.rouhe-interpretive.v6as the last versioned lineage artefact - Promoted
Modelfile-rouheas the single canonical active Modelfile - Normalized Modelfile naming for Ollama list clarity
- No behavioral or architectural changes
- Serves as the pre-LoRA baseline checkpoint
v1.3.1 — 2026-01-03
- Added and locked
ROUHE_ARCHITECTURE.mdas the authoritative, platform-agnostic behavioral specification - Formalized Rouhe-Core, Behavioral Layers, Guardrails, Failure Modes, and Non-Goals
- Established explicit rejection of assistant-style and vending-machine interaction
- No functional, scope, or runtime changes
v1.3.0 — 2026-01-03
- Repository re-scoped as a local-only reference demo
- Renamed under RIGHTBUSINESS as rb-local-llm-persona-demo
- Root README replaced to enforce non-product scope
- Original persona artefacts preserved under /archive
v1.2.0 — 2026-01-03
- Ownership transferred to RIGHTBUSINESS PTE. LTD.
- Updated LICENSE and legal notices
v1.1.0 — 2025-06-10
- Hardened restrictions: no code generation, no content writing
- File analysis strictly non-creative
v1.0.0 — 2025-06-08
- Initial release of Dr. Greyson Rouhe persona artefact
- Full persona logic, tone structure, and licensing defined