AI Operations

Infrastructure cost monitoring & model efficiency

Token Dashboard 787 sessions tracked

Sessions (7d)

14

Cache Hit Rate

70.8%

Avg Duration

19m

Avg Input Tok

6,654,974

Avg Output Tok

77,119

Total Sessions

787

since 2026-02-17

Model Mix (7-day)

opus
100.0%

Session Quality (7-day)

9

Excellent

85+

5

Good

70-84

0

Warning

50-69

0

Poor

<50

Daily Sessions & Cache Hit (14-day)

Weekly Overhead Trend

2026-W18
846,724
1s
2026-W19
4,473,303
22s
2026-W20
5,592,477
19s
2026-W21
11,338,891
19s
2026-W22
7,988,556
8s
2026-W23
6,744,114
21s
2026-W24
7,764,137
12s

Avg input tokens per session. Rising = growing CLAUDE.md / memory overhead.

Project Breakdown (7-day)

Project Sessions Avg Min Input Tok
rgdm-agency 14 19 93,169,649

Skill Metrics (from Mission Control)

ai cost pacing pct

83.00

ai cost rolling avg 7d

36.30

cache hit rate avg

0.90

model waste estimate

81.14

opus ratio

0.92

overhead avg input tokens

8,022,192.00

overhead wow change pct

-37.30

session quality avg

87.10