OBSIDIAN_TODO.md

Obsidian & 3D Workflow Integration Todo List

This document tracks the progress and remaining tasks for integrating Obsidian Canvas and 3D spatial reasoning into the Pipecat workflow engine.

Phase 1: Core Schema & Converter (In Progress)

• Create Design Document: docs/OBSIDIAN_WORKFLOW_DESIGN.md
• Update Workflow Schema: Added position (x,y,z), dimensions (w,h,d), style to Node class in pipecatapp/workflow/node.py.
• Implement Canvas Converter: Created pipecatapp/workflow/canvas_converter.py for bidirectional conversion.
• Verify Conversion: Write tests to ensure data integrity during round-trip conversion.

Phase 2: 3D Visualization (Next Steps)

• Select 3D Library: Decide between Three.js (standard web) or A-Frame (VR-first). Recommendation: Three.js for broad compatibility, wrapped in A-Frame for VR.
• Create 3D Viewer: Implement pipecatapp/static/workflow_3d.html.
  • Render nodes as 3D objects (boxes/spheres) at correct (x,y,z).
  • Draw connections as 3D lines/curves.
  • Implement camera controls (Orbit, Pan, Zoom).
• Implement "Semantic Zoom":
  • Define Z-axis layers (e.g., Z=0 Logic, Z=100 Summary).
  • Add UI slider or scroll interaction to fade layers in/out based on camera depth.

Phase 3: The "Gardener" Agent (Automation)

• File Watcher Service: Create a background service that monitors the Obsidian Vault path.
• Active Document Logic:
  • Detect "seeds" (new files/nodes with specific tags).
  • Parse directives (e.g., <!-- run: python -->).
  • Trigger Workflow execution.
  • Write results back to the file (Markdown append or Canvas node update).

Phase 4: Advanced Canvas Features

• Scope/Group Support: Map Canvas "Groups" to Workflow "Scopes" or "Sub-workflows".
• Rich Content Embedding: Allow Workflow nodes to reference local assets (images, PDFs) from the Vault.
• Bi-directional Sync: Real-time updates (using WebSocket) to the open Canvas file in Obsidian (if possible via plugin) or just file polling.