Agent Coordination

Agent Coordination Framework

Overview

This skill provides the reference documentation for multi-agent coordination, including communication protocols, workflow standards, and collaboration patterns.

Mandatory Reporting Protocol

All specialized work MUST follow this communication protocol:

1. Before Work (Authorization Request)

"Requesting authorization for [task]"
"Scope: [detailed scope]"
"Dependencies: [other skills/tasks]"
"Estimated time: [duration]"

2. During Work (Progress Updates)

Required every 30-60 minutes:

"Progress update"
"Completed: [specific accomplishments]"
"Current: [what's in progress]"
"Next: [next steps]"
"Blockers: [any issues]"
"Commits: [commits made since last update]"

3. After Work (Completion Report)

"Completed [task]"
"Deliverables: [what was produced]"
"Files changed: [list of files]"
"Commits made: [list of commits]"
"Tests: [test status]"
"Ready for: [next steps]"

Skill Ecosystem

Skill Capabilities

Coordination Patterns

Sequential Workflow

Planning → Research → Backend → UI → Testing → Documentation → Version Control

Parallel Opportunities

Research + Planning (parallel)
Backend + early UI prototyping (parallel)
Testing + Documentation (parallel)

Commit Frequency Protocol

Mandatory Intervals

• Maximum: 60 minutes between commits
• Recommended: 30-45 minutes for active work
• Required triggers:
  • After each component completion
  • After tests added/updated
  • After integration work
  • After documentation updates
  • Before requesting review

Commit Message Format

git commit -m "$(cat <<'EOF'
type(scope): description

- Detailed changes
- List modifications

Co-Authored-By: Claude <noreply@anthropic.com>
EOF
)"

Quality Gates

Phase Transitions

Before progressing between workflow phases:

1. Documentation complete and accurate
2. Types properly defined
3. Tests written and passing
4. Tech stack compliance verified
5. Error handling implemented

Approval Requirements

• Standard tasks: Skill self-validates
• Complex tasks: Orchestration approval required
• Critical changes: User approval required

Complexity Assessment

Scoring System

• Simple file edit: 1
• Single component creation: 2
• Data model creation: 3
• Service implementation: 4
• UI component development: 5
• Testing integration: 6
• Multi-component feature: 8
• Full feature with backend and UI: 9
• System architecture changes: 10

Orchestration Triggers

• IF complexity_score >= 7 THEN orchestration_required
• IF multiple_skills_needed THEN coordination_required
• IF backend_AND_frontend THEN orchestration_required
• IF testing_mentioned THEN testing_skill_activated

Conflict Prevention

Branch Isolation

• Each feature on dedicated branch
• Clear ownership during development phases
• Scheduled merge points
• Independent test suites per feature

File Ownership

• Explicit ownership during active work
• No concurrent edits to same files
• Clear handoff protocols

Emergency Protocols

Work Failure Recovery

1. Immediate pause all related work
2. Root cause analysis
3. Rapid replanning
4. Orchestration review
5. Coordinated resume

Conflict Resolution

1. Conflict detection
2. Immediate escalation
3. Work stream isolation
4. Resolution planning
5. Validated resume

Success Metrics

• Zero Conflicts: No merge conflicts between parallel work
• 100% Coordination: All skills working from synchronized context
• >99% Approval: All work passes quality review
• Measurable Outcomes: Every task has specific success metrics
• Complete Documentation: All deliverables properly documented