Sutra

# Sutra Roadmap: From Library to Engine **Current Status**: `v0.1.0` (Feature Complete, Secure, Polished) **Target Status**: `v0.2.0` (Dynamic Context Orchestration) --- ## Phase 1: The Cognitive Foundry (Thinking Models) - [x] **Refactor Directory Structure** - [x] **Implement `cognitive/thinking_models.py`** - [x] `understand_question` - [x] `backtracking` - [x] `symbolic_abstract` - [x] `verify_logic` - [x] **Implement Prompt Program Generator (`core/programs.py`)** - [x] `math` program type. - [x] `debate` program type. ## Phase 2: The Cell Laboratory (Memory & State) - [x] **Implement `core/cells.py`** - [x] `cell.protocol.key_value` - [x] `cell.protocol.windowed` - [x] `cell.protocol.episodic` ## Phase 3: The Organ Assembly (Orchestration) - [x] **Implement `systems/organs.py`** - [x] `organ.research_synthesis` (Scout -> Architect -> Scribe) - [x] `organ.debate_council` ## Phase 4: Maintenance & Quality - [x] **Add Schema Validators** - [x] Pydantic models for all Tool inputs. - [x] **Testing & Linting** - [x] 100% test pass rate. - [x] Ruff linting clean. - [ ] **Documentation** - [ ] Auto-generate API reference. ## Phase 5: v0.2.0 Prospects - [ ] **Serena/Indexing Integration** - Reintroduce optional Serena indexing to keep abstraction graph and prefs accessible. - [ ] **Dynamic Orchestration** - Agents that can self-select and chain these tools.