agent-performance

# Agent Performance Dashboard

## Purpose

Provides visibility into agent usage patterns to optimize delegation and identify improvement opportunities.

## When I Activate

I automatically load when you mention:

- "agent performance" or "agent metrics"
- "agent dashboard" or "agent usage"
- "which agents are used" or "underutilized agents"
- "agent success rate" or "agent statistics"

## What I Do

1. **Track Invocations**: Record agent usage via workflow tracker
2. **Measure Success**: Track completion rates per agent
3. **Analyze Patterns**: Identify usage trends and gaps
4. **Generate Reports**: Create actionable dashboards

## Quick Start

```
User: "Show me agent performance metrics"
Skill: *activates automatically*
       "Generating agent performance report..."
```

## Core Capabilities

### 1. Report Generation

Generate a performance report by reading workflow logs and aggregating agent metrics:

```
User: "Generate agent performance report"
```

Report includes:

- Invocation counts per agent
- Success/failure rates
- Average completion times (when tracked)
- Underutilized agents list
- Recommendations for optimization

### 2. Live Tracking

Track agent invocations during workflow execution using the existing `workflow_tracker`:

```python
# Already available in .claude/tools/amplihack/hooks/workflow_tracker.py
from workflow_tracker import log_agent_invocation

log_agent_invocation(
    agent_name="architect",
    purpose="Design authentication module",
    step_number=2
)
```

### 3. Metrics Storage

Metrics are stored in:

- **Raw logs**: `.claude/runtime/logs/workflow_adherence/workflow_execution.jsonl`
- **Aggregated**: `.claude/runtime/metrics/agent_performance.yaml`

## Report Format

### Summary Dashboard

```yaml
# Agent Performance Summary
# Generated: 2025-11-25

total_invocations: 142

agents:
  architect:
    invocations: 45
    success_rate: 95.6%
    avg_duration_ms: 2340
    trend: increasing

  builder:
    invocations: 38
    success_rate