My Dashboard

37 finalized PRs in past 30d

Output Quality
Avg Post-Open Commits
1.6
↓ 0.3 vs prior 30d

How many commits were pushed after the PR was opened — fewer means the agent got it right the first time.

CI Success Rate
82%
↑ 4% vs prior 30d

How often agent-generated code passes CI on the first try — low rates mean the agent is producing code that doesn't build or pass tests.

Avg Line Revisit Rate
0.16
↑ 0.01 vs prior 30d

How often the same files get changed across multiple recent PRs — high churn suggests the agent isn't making durable changes.

Prompt Efficiency
Avg Iteration Depth
8
↓ 1 vs prior 30d

How many back-and-forth turns it takes to finish a task — fewer means your prompts are clear and the agent stays on track.

Avg Token Cost
$1.82
↑ $0.17 vs prior 30d

How much you spent on AI tokens to produce this PR — tracks whether the agent is cost-efficient or burning through tokens.

Avg Cache Hit Rate
65%
↑ 5% vs prior 30d

How much of the agent's input was served from cache — higher means you're spending less on repeated context.

Agent Behavior
Avg Sidechain Rate
12%
↓ 3% vs prior 30d

How often the agent backtracks down dead-end reasoning paths — lower means less wasted work and faster completions.

Avg Re-Read Rate
1.66
↓ 0.11 vs prior 30d

How often the agent re-reads files it already opened — excessive re-reads waste tokens and slow things down.

Avg Autonomy Score
7.7
↑ 0.8 vs prior 30d

How much work the agent does between human interventions — higher means it operates independently with less hand-holding.

View all my pull requests →