Skip to content

Latest commit

 

History

History
97 lines (78 loc) · 2.96 KB

PHASE-3-ASSIGNMENTS.md

File metadata and controls

97 lines (78 loc) · 2.96 KB

Phase 3: Performance and Quality - Agent Assignments

⚠️ MANDATORY: Read Before Starting

  1. READ CLAUDE.md - Top section about AGENT-GUIDELINES.md
  2. READ AGENT-GUIDELINES.md - Especially "Parallel Development Workflow" section
  3. FOLLOW ALL GUIDELINES - Show command outputs, create branches, run scripts

Agent Assignments

Agent 1: Performance Benchmarks

Branch: feature/phase3-benchmarks Files to create/modify:

  • test/Tethys.Benchmarks/Tethys.Benchmarks.csproj (new project)
  • test/Tethys.Benchmarks/ResultBenchmarks.cs
  • test/Tethys.Benchmarks/FunctionalMethodsBenchmarks.cs

Tasks:

  1. Follow AGENT-GUIDELINES.md "Parallel Development Workflow"
  2. Create benchmark project using BenchmarkDotNet
  3. Implement benchmarks for:
    • Result creation (success vs failure)
    • Match method performance
    • Map/FlatMap operations
    • Callback methods
    • Try pattern overhead
  4. Run benchmarks and document results

Agent 2: Code Coverage Setup

Branch: feature/phase3-coverage Files to modify:

  • .github/workflows/quality-gates.yml
  • Directory.Build.props (if needed)
  • Update all .csproj files for coverage

Tasks:

  1. Follow AGENT-GUIDELINES.md "Parallel Development Workflow"
  2. Set up Coverlet for code coverage
  3. Update CI/CD to include coverage reports
  4. Add coverage badges to README
  5. Ensure coverage gates work in PR checks

Agent 3: Performance Optimizations

Branch: feature/phase3-optimization Files to modify:

  • src/Tethys.Results/Result.cs (Combine method)
  • src/Tethys.Results/GenericResult.cs (Combine method)
  • src/Tethys.Results/ResultExtensions.cs (async methods)

Tasks:

  1. Follow AGENT-GUIDELINES.md "Parallel Development Workflow"
  2. Run Agent 1's benchmarks first (baseline)
  3. Optimize Combine methods (avoid multiple enumerations)
  4. Add ConfigureAwait(false) to all async methods
  5. Re-run benchmarks to show improvements

Coordination Rules

  1. All agents start simultaneously

  2. All agents MUST:

    • Show branch creation: git checkout -b feature/phase3-XXX
    • Create progress files: progress/agent-N-status.md
    • Show test/build outputs
    • Run verification scripts and show output
    • Create completion marker: progress/READY-agent-N

  3. Merge Order:

    • Agent 2 first (infrastructure)
    • Agent 1 second (baseline metrics)
    • Agent 3 last (shows improvements)

Example of Expected Output

Every agent response should look like:

## Agent 1: Performance Benchmarks

### Branch Setup
```bash
git checkout -b feature/phase3-benchmarks
# OUTPUT: Switched to a new branch 'feature/phase3-benchmarks'

git branch --show-current
# OUTPUT: feature/phase3-benchmarks

[... rest of work following AGENT-GUIDELINES.md format ...]


## Verification Before Merge

No agent work is complete until:
1. Branch created and shown
2. All verification scripts pass with output shown
3. Progress files exist with completion marker
4. All tests pass
5. Coverage maintained >95%