ShadowHound TODO List¶

Last Updated: 2025-10-13 (Major revision based on status analysis)
Current Phase: Phase 0 (Infrastructure) → Phase 1 (Skills Foundation)

Key Changes: - Realistic assessment: We're completing Phase 0, not Phase 1 - Focus on actual implementation (skills, testing) - Leverage cloud agents for suitable tasks - Clear acceptance criteria for all tasks

See Status Analysis (Oct 2025) for detailed assessment.

🔴 Phase 0: Complete Infrastructure (10% Remaining)¶

Goal: Finish testing infrastructure before moving to Phase 1

Testing Infrastructure ⭐¶

[ ] Set up pytest configuration
Create pytest.ini in workspace root
Configure test discovery patterns
Set up coverage reporting
Acceptance: pytest runs successfully from workspace root
Related: All package test/ directories
[ ] Create mock robot interface
Mock DIMOS API calls
Configurable behavior for testing
State tracking for assertions
Acceptance: Can test skills without real robot
Related: shadowhound_skills/shadowhound_skills/mock_robot.py
[ ] Add CI for testing
GitHub Actions workflow for tests
Run on every PR
Block merge if tests fail
Acceptance: CI runs tests automatically
Related: .github/workflows/tests.yml

Code Quality in CI¶

[ ] Add black/isort/flake8 checks
Format checking (not auto-format)
Import sorting validation
Linting with project rules
Acceptance: CI fails on style violations
Related: .github/workflows/quality.yml

Estimated Time: 1-2 days
Cloud Agent Candidate: ✅ Yes - Well-defined setup tasks

🔴 Phase 1: Skills Foundation (Priority Tasks)¶

Goal: Implement core skills infrastructure + 3-5 working skills

See Roadmap Phase 1 for complete deliverables.

Core Infrastructure (Week 1)¶

RobotInterface Implementation ⭐⭐⭐¶

[ ] Design RobotInterface class
Define interface methods (move, rotate, stop, etc.)
DIMOS API integration points
Safety parameter configuration
Acceptance: Interface design documented
Related: docs/software/robot_interface_design.md (create)
[ ] Implement RobotInterface
Connect to DIMOS API
Wrap basic motion commands
Add safety clamps (velocity, bounds)
Timeout and error handling
Acceptance: Can control robot via interface
Related: shadowhound_skills/shadowhound_skills/robot_interface.py
[ ] Test RobotInterface
Unit tests with mock DIMOS
Integration tests with real/sim robot
Safety parameter validation
Acceptance: >80% test coverage
Related: shadowhound_skills/test/test_robot_interface.py

SkillRegistry Implementation ⭐⭐⭐¶

[ ] Create Skill base class
Abstract validate_params() method
Abstract execute() method
Common telemetry collection
Acceptance: Base class usable for all skills
Related: shadowhound_skills/shadowhound_skills/skill_base.py
[ ] Implement SkillRegistry
@register_skill decorator
Skill discovery mechanism
execute(skill_name, **params) method
Validation before execution
Acceptance: Can register and execute skills
Related: shadowhound_skills/shadowhound_skills/skill_registry.py
[ ] Create SkillResult dataclass
success: bool
data: dict
error: Optional[str]
telemetry: dict
Acceptance: All skills return SkillResult
Related: shadowhound_skills/shadowhound_skills/skill_base.py

First 3 Skills (Week 1-2)¶

Skill 1: nav.stop¶

[ ] Implement StopSkill
Immediately set all velocities to zero
No parameters needed
Always succeeds (unless comms failure)
Acceptance: Robot stops within 0.5s
Related: shadowhound_skills/shadowhound_skills/skills/navigation.py
[ ] Test StopSkill
Unit test with mock
Integration test (verify velocities zero)
Acceptance: Tests pass
Related: shadowhound_skills/test/test_navigation_skills.py

Skill 2: nav.rotate¶

[ ] Implement RotateSkill
Parameter: yaw (degrees, -180 to 180)
Rotate robot by specified angle
Timeout after 10s
Acceptance: Rotates accurately ±5 degrees
Related: shadowhound_skills/shadowhound_skills/skills/navigation.py
[ ] Test RotateSkill
Unit tests (parameter validation, timeout)
Integration test (measure actual rotation)
Acceptance: Tests pass
Related: shadowhound_skills/test/test_navigation_skills.py

Skill 3: report.log¶

[ ] Implement LogSkill
Parameter: message (string, 1-500 chars)
Log to ROS logger with timestamp
Always succeeds
Acceptance: Message appears in logs
Related: shadowhound_skills/shadowhound_skills/skills/reporting.py
[ ] Test LogSkill
Unit test (parameter validation)
Integration test (verify log output)
Acceptance: Tests pass
Related: shadowhound_skills/test/test_reporting_skills.py

Documentation (Week 2)¶

[ ] Document skills API
How to implement a skill
How to register a skill
How to execute a skill
Acceptance: Developer can add new skill
Related: docs/software/skills_api_guide.md (create)
[ ] Document implemented skills
Parameters, return values, examples
Success/failure scenarios
Performance characteristics
Acceptance: Each skill has complete docs
Related: docs/software/skills_reference.md (create)

Estimated Time: 2-3 weeks total for Phase 1 core
Cloud Agent Candidates: - ✅ Testing infrastructure (skill 1-3 after examples set) - ✅ Documentation generation - ⚠️ RobotInterface (needs domain knowledge)

🟡 Phase 1: Additional Skills (After Core)¶

Prerequisite: Core infrastructure + first 3 skills working

[ ] nav.translate - Move forward/backward
[ ] nav.goto - Navigate to pose (x, y, yaw)
[ ] nav.follow_path - Follow waypoint list

Perception Skills (3)¶

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

System Skills (2-3)¶

[ ] system.health_check - Verify systems operational
[ ] system.emergency_stop - Safe emergency shutdown
[ ] system.battery_status - Report battery level

Estimated Time: 1-2 weeks
Cloud Agent Candidate: ✅ Yes - Follow established patterns

🟢 Phase 2: Mission Agent (Future)¶

Prerequisite: Phase 1 complete (10-15 working skills)

See Roadmap Phase 2 for details.

High Priority (When Phase 2 Starts)¶

[ ] DIMOS agent integration
[ ] Mission executor node
[ ] Web dashboard UI ← Excellent cloud agent task
[ ] Camera feed integration
[ ] Example missions (5+)

✅ Recently Completed¶

October 2025 (Week 2) - Infrastructure & Documentation¶

✅ Documentation pipeline reversal (Issue #20)
Author in standard Markdown, generate Obsidian vault locally
Created tools/obsidian_convert.py and scripts/generate_obsidian_vault.sh
Simplified CI, removed 165k lines of build artifacts
8x velocity improvement demonstrated
✅ Cloud agent collaboration workflow
Comprehensive workflow documentation (600+ lines)
Quick start guide (200+ lines)
GitHub issue templates (cloud agent, feature, bug)
Integrated into development hub
✅ Submodule protection
.gitattributes marking (linguist-vendored, -diff)
VS Code formatter exclusions
Troubleshooting documentation
✅ Project status analysis
Comprehensive assessment of actual vs. claimed state
Lessons learned documentation
Roadmap revision with specific deliverables
TODO reorganization by phase

October 2025 (Week 1) - Documentation Cleanup¶

✅ Fixed documentation navigation and structure
✅ Created hub pages for all major categories
✅ Optimized Obsidian graph view with colors
✅ Eliminated phantom nodes and broken links
✅ Applied YAML front-matter consistently
✅ Created wikilink validation tool

September - October 2025 - Bootstrap Phase¶

✅ Devcontainer with ROS2 Humble + DIMOS
✅ Package structure (3 packages: bringup, mission_agent, skills)
✅ Build system working (colcon build)
✅ DIMOS submodule integration
✅ Helpful aliases (cb, source-ws, etc.)

Pre-October 2025 (Historical Context)¶

Note: These were marked as complete in previous versions but may not actually exist in codebase. See Status Analysis for reality check.

⚠️ Local LLM integration (claimed, needs verification)
⚠️ Agent improvements for tool calling (claimed, needs verification)
⚠️ Camera feed fixes (claimed, needs verification)
⚠️ Web UI optimizations (claimed, needs verification)
⚠️ Multi-step execution (claimed, needs verification)

📋 Task Management¶

Priority Levels¶

🔴 Phase 0 (Current): Blocking Phase 1 start
🔴 Phase 1 (Next): Core skills implementation
🟡 Phase 1 (Additional): Extra skills after core
🟢 Phase 2 (Future): Mission agent work
🔵 Future Phases: Phase 3-4 work

Adding New Tasks¶

When adding tasks, include: 1. Clear title: What needs to be done 2. Context/Why: Reason it's needed 3. Acceptance criteria: How to verify completion 4. Related files/docs: Links to relevant code or docs 5. Cloud agent candidate: ✅/⚠️/❌ suitable for cloud agent?

Example:

- [ ] Implement RotateSkill
  - Context: Basic navigation skill for Phase 1
  - Acceptance: Rotates accurately ±5 degrees within 10s
  - Related: `shadowhound_skills/shadowhound_skills/skills/navigation.py`
  - Cloud Agent: ⚠️ After first skill as example

Cloud Agent Strategy¶

✅ Good candidates: Well-defined, repeatable, testable
⚠️ Maybe: After examples established
❌ Not suitable: Requires domain knowledge, architectural decisions

See: Cloud Agent Workflow for details

📊 Progress Tracking¶

Current Sprint (Phase 0 → Phase 1)¶

Target: Complete Phase 0, start Phase 1 core
Duration: 1-2 weeks
Status: Phase 0 at 90%

Completion Metrics¶

Phase 0: 90% (9/10 tasks complete)
Phase 1 Core: 0% (0/15 tasks started)
Phase 1 Additional: 0% (0/9 tasks started)

Next Milestone¶

Phase 1 Core Complete: - All infrastructure tasks done - First 3 skills working and tested - Documentation complete - Target: November 15, 2025