| Agent | Actions ↓ | Success | Avg Time | Flags | Escalations |
|---|---|---|---|---|---|
Ma Maren Chief of Staff | 152 | 97.2% | 0.3h | 2 | 0 |
Br Brue Quality Critic | 102 | 94.5% | 0.8h | 87 | 4 |
No Nova Orchestrator | 60 | 99.1% | 0.1h | 0 | 0 |
CC Claude Code Executor | 52 | 96.8% | 4.2h | 0 | 1 |
Tu Tuck Knowledge Keeper | 36 | 98.5% | 2.1h | 0 | 0 |
Maren's approval rate dropped 12% last week
Spec rejection rate increased from 3% to 15%. Root cause: 4 specs submitted without initiative validation step. Maren is re-running all 4 through the full checklist.
Brue flagged 3 identical PAT-004 patterns today
PAT-FEAT-004 (missing loading state) triggered 3 times on 3 different PRs. May indicate a systemic gap in component templates rather than individual oversights.
Claude Code avg task completion time improved 18%
Average time from in_progress to shipped decreased from 5.1h to 4.2h over the last 30 days. Likely due to improved context packs reducing initial exploration time.
KB consolidation queue growing — 8 items pending
Tuck's consolidation backlog grew from 2 to 8 items this week. Three session summaries from S205-S207 and 5 duplicate ADR cross-references awaiting merge.
| Model | Provider | Tasks | Success | Avg Cost |
|---|---|---|---|---|
| Claude 4.6 Opus | Anthropic | 186 | 97.3% | $3.42 |
| Claude 4.6 Sonnet | Anthropic | 420 | 96.9% | $0.84 |
| GPT-5 | OpenAI | 98 | 95.1% | $2.90 |
| GPT-4o-mini | OpenAI | 1240 | 94.2% | $0.12 |
| Gemini 2.5 Pro | 45 | 93.8% | $1.65 |