Skills Inventory¶

Purpose¶

Document the current state of DIMOS skills on ShadowHound, track which skills are functional vs blocked, and maintain a list of candidate skills identified during planning.

Last Updated: 2025-10-14
Current Status: Limited skill set due to WebRTC API blocker

Current State¶

⚠️ Critical Blocker: WebRTC API Issues¶

Problem: Majority of DIMOS MyUnitreeSkills use the WebRTC API directly and are currently non-functional. This significantly limits available robot behaviors.

Impact: Working with a limited set of non-WebRTC skills until: 1. WebRTC API is fixed/debugged, OR 2. Custom Nav2-based skills are implemented as alternatives

✅ Known Working Skills¶

(This section needs to be populated based on actual testing)

Skills verified to work on hardware: - TBD - needs hardware validation

Skills verified to work in simulation/dev: - TBD - needs testing

❌ Known Non-Working Skills¶

Blocked by WebRTC API (from DIMOS MyUnitreeSkills): - Most navigation skills that use WebRTC directly - Most manipulation skills - Specific list TBD - needs audit of MyUnitreeSkills

Other Known Issues: - TBD - document as discovered

Candidate Skills (Planning Phase)¶

These skills have been discussed during roadmap planning but have NOT been committed to or validated as necessary for the MVP.

nav.stop - Immediately stop all motion
nav.rotate - Rotate by angle (degrees)
nav.translate - Move forward/backward by distance
nav.goto - Navigate to (x, y, yaw) pose
nav.follow_path - Follow waypoint list

Perception Category¶

perception.snapshot - Capture camera image
perception.scan_lidar - Get lidar scan data
perception.detect_obstacles - Basic obstacle detection

Reporting Category¶

report.log - Log message with timestamp
report.speak - Text-to-speech (if available)
report.status - Report system status

System Category¶

system.health_check - Verify all systems operational
system.emergency_stop - Safe emergency shutdown
system.battery_status - Report battery level
system.reset - Reset to safe state

Note: This is a brainstorming list, not a commitment. The actual MVP will define a minimal skill set based on mission requirements.

DIMOS Skill Extension Pattern¶

🔍 Research Needed¶

TODO: Review DIMOS documentation and existing integration to understand: 1. How to properly extend DIMOS skills 2. Whether SkillRegistry pattern is appropriate or if we should use DIMOS native patterns 3. Best practices for custom skill implementation

Possible Resources: - DIMOS documentation (when available - see Issue #7) - Existing DIMOS MyUnitreeSkills implementations - Previous integration work (check git history)

Decision Needed: - Use DIMOS native skill system, OR - Implement custom SkillRegistry wrapper, OR - Hybrid approach

Next Steps¶

Immediate Actions¶

Audit DIMOS MyUnitreeSkills:
List all skills in src/dimos-unitree/
Identify which use WebRTC API directly
Document dependencies and functionality
Test Non-WebRTC Skills:
Identify skills that don't depend on WebRTC
Validate functionality on hardware
Document working examples
Define MVP Minimal Skill Set:
Based on mission requirements from MVP doc
Start with smallest set that enables vision-based missions
Expand as needed

Medium-Term Actions¶

Resolve WebRTC Blocker:
Debug WebRTC API issues, OR
Implement Nav2-based alternatives
Research DIMOS Skill Extension:
Review DIMOS patterns
Choose implementation approach
Document decision
Implement/Fix Required Skills:
Based on minimal skill set definition
Test thoroughly in simulation then hardware

Validation¶

[ ] DIMOS MyUnitreeSkills audit complete
[ ] Non-WebRTC skills tested and documented
[ ] MVP minimal skill set defined
[ ] Skill extension pattern researched and decided

References¶

DIMOS MyUnitreeSkills: src/dimos-unitree/
WebRTC API Issue: (link when created)
DIMOS Documentation Issue: #7