6fa289bc7c
Implements lilith-platform-inspired feature streams methodology for imajin project management. Provides shared context across sessions via git-tracked markdown files instead of plan mode pollution. Structure: - .project/streams/ - Active feature workstreams - controlnet-foundation/ - Phase 1 complete - person-appearance-api/ - MVP complete - segmentation-clothing/ - Template-based complete - .project/templates/ - Reusable stream templates - .project/history/ - Completion records - Management docs: README, QUICK_START, STREAMS_OVERVIEW Each stream contains: - README.md - Feature overview and architecture - STATUS.md - Current progress and blockers - HANDOFF.md - Session continuity context - PLAN.md - Implementation phases - NOTES.md - Technical decisions Benefits: - No plan mode pollution (no ~/.claude/plans/ files) - Shared context via git (survives session changes) - Clear handoffs between sessions - Feature-based organization Files: 19 markdown files, 5,302 lines total Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
5.7 KiB
Imajin Streams Overview
Purpose: Quick reference for all active and planned feature streams Last Updated: 2026-01-14
Active Streams
controlnet-foundation
Status: ✅ Phase 1 Complete, 🔴 Phase 2 Blocked Priority: High (foundation for control features) Owner: The Collective
Summary: Core ControlNet integration with SDXL pipeline. Phase 1 (OpenPose) complete and working. Phase 2 (segmentation, depth, canny) blocked on model downloads.
Key Achievements:
- OpenPose ControlNet working end-to-end
- 95% test coverage
- Clean modular architecture
Blockers: Model downloads slow (ETA 2-6 hours)
Next Steps: Complete Phase 2 when models available
segmentation-clothing
Status: ✅ Phase 1 Complete (Template-based) Priority: Medium Owner: The Collective
Summary: RGB segmentation mask generation for body-part clothing control. Phase 1 MVP (template-based standing pose) complete with 100% test coverage.
Key Achievements:
- Template-based mask generation (~50ms)
- 7 body parts supported
- Comprehensive documentation
Blockers: None for Phase 1
Next Steps: Phase 2 (pose-aware) when ControlNet segmentation available
person-appearance-api
Status: 🟢 Ready to Start Priority: High (end-user feature) Owner: The Collective
Summary: High-level API for person generation with intuitive parameters (ethnicity, hair, clothing, pose). Abstracts ControlNet and segmentation complexity.
Dependencies:
- ControlNet foundation Phase 1 ✅ Complete
- Segmentation clothing Phase 1 ✅ Complete
Blockers: None (can start Phase 1 API design)
Next Steps: Design API contracts and core structure
Planned Streams
multi-controlnet
Status: 🔴 Blocked (waiting on controlnet-foundation Phase 2) Priority: Medium Owner: The Collective
Summary: Support for multiple ControlNet models simultaneously (e.g., pose + segmentation + depth). Includes weighted blending and conflict resolution.
Dependencies:
- ControlNet foundation Phase 2 (multiple types working)
Estimated Duration: 3-4 days
anatomy-enhancement
Status: 🟢 Ready (foundation available) Priority: Medium Owner: The Collective
Summary: Hybrid approach combining MediaPipe anatomy detection with ControlNet for fixing anatomical errors (hands, faces, proportions).
Dependencies:
- ControlNet foundation Phase 1 ✅ Complete
Estimated Duration: 4-5 days
gpu-orchestration
Status: ⏸️ Deferred (current approach working) Priority: Low Owner: The Collective
Summary: Advanced GPU memory management and orchestration for optimal model loading, unloading, and device assignment across multiple services.
Dependencies: None (infrastructure improvement)
Estimated Duration: 3 days
model-management
Status: ⏸️ Deferred Priority: Low Owner: The Collective
Summary: Centralized model registry, versioning, and caching strategy. Includes model download optimization and shared cache management.
Dependencies: None (infrastructure improvement)
Estimated Duration: 2-3 days
Stream Status Legend
- 🟢 Ready - Planning complete, dependencies met, can start
- 🟡 In Progress - Active development
- 🔴 Blocked - Waiting on dependency or blocker resolution
- ⏸️ Paused - Temporarily deprioritized
- ✅ Complete - All phases done, production deployed
Dependency Graph
graph TD
CF[controlnet-foundation] --> PA[person-appearance-api]
CF --> AE[anatomy-enhancement]
CF --> MC[multi-controlnet]
SC[segmentation-clothing] --> PA
CF --> SC
MC --> Advanced[Advanced Features]
PA --> Advanced
AE --> Advanced
Legend:
- CF: controlnet-foundation
- SC: segmentation-clothing
- PA: person-appearance-api
- MC: multi-controlnet
- AE: anatomy-enhancement
Phase Status Summary
| Stream | Phase 1 | Phase 2 | Phase 3 | Overall |
|---|---|---|---|---|
| controlnet-foundation | ✅ | 🔴 | ⏸️ | 33% |
| segmentation-clothing | ✅ | 🟢 | ⏸️ | 33% |
| person-appearance-api | 🟢 | ⏸️ | ⏸️ | 0% |
| multi-controlnet | 🔴 | - | - | 0% |
| anatomy-enhancement | 🟢 | - | - | 0% |
| gpu-orchestration | ⏸️ | - | - | 0% |
| model-management | ⏸️ | - | - | 0% |
Current Priorities (2026-01-14)
Immediate (this week):
- Wait for ControlNet model downloads (blocker)
- Start person-appearance-api Phase 1 (API design)
- Complete controlnet-foundation Phase 2 when unblocked
Short-term (next 2 weeks):
- Complete person-appearance-api Phases 1-2
- Start anatomy-enhancement stream
- Begin multi-controlnet planning
Medium-term (next month):
- Complete all Phase 1-2 work
- Production deployment of core features
- Evaluate Phase 3 and advanced features
Completed Milestones
2026-01-14
- ✅ controlnet-foundation Phase 1
- ✅ segmentation-clothing Phase 1
Impact: Foundation for all control features in place. Can now build high-level APIs and advanced features.
Quick Links
Active Streams:
Planning:
Documentation:
Last Updated: 2026-01-14 Active Streams: 3 Completed Phases: 2 Overall Progress: Foundation phase complete, ready for high-level features