subagent-driven-development

# Subagent-Driven Development (Shannon-Enhanced)

## Overview

**Execute plans with fresh subagent per task + quantitative quality gates.**

Each task: Subagent implementation → Code review → Quality score → Advance if > 0.80.

Shannon enhancement: Numerical quality scoring, pattern learning from task history, MCP work tracking.

## When to Use

**Same-session task execution:**
- Staying in this session
- Tasks mostly independent
- Want continuous progress with quality gates
- Plan already written and validated

**Don't use when:**
- Need to review plan first
- Tasks are tightly coupled
- Plan needs revision

## Quality Scoring (0.00-1.00)

**Per-task scoring:**
```
task_quality_score = (
  (implementation_completeness) * 0.4 +
  (test_coverage) * 0.35 +
  (code_review_score) * 0.25
)
```

**Thresholds:**
- 0.00-0.50: Needs major rework (Critical issues)
- 0.51-0.80: Has issues, fixable (Important issues)
- 0.81-0.95: Good, minor cleanup (Minor issues)
- 0.96-1.00: Excellent, ready to advance

**Gate rule:** Only advance to next task if task_quality_score > 0.80

## The Process

### 1. Load Plan & Setup Tracking

```bash
# Serena metric initialization
plan_metadata = {
  plan_file: "path/to/plan.md",
  total_tasks: N,
  tasks_completed: 0,
  cumulative_quality_score: 0.0,
  estimated_completion_time: "X hours"
}
```

Create TodoWrite with all tasks, all in "pending" state.

### 2. Execute Task (Fresh Subagent)

Mark task as in_progress. Dispatch:

```markdown
**Task N:** [task_name]
Implementation scope: [read task N from plan]

Requirements:
1. Implement exactly what task specifies
2. Write tests (TDD if task requires)
3. Verify all tests pass
4. Commit with conventional message
5. Report: what you implemented, test results, files changed

Work directory: [path]
```

Subagent reports implementation summary.

### 3. Code Review Gate

**Dispatch code-reviewer with metrics:**

```bash
# Review metrics (Serena)
review_context = {
  task_number: N,
  what_implemented: "