RGDM OS

AI Operations

Infrastructure cost monitoring & model efficiency

Token Dashboard 670 sessions tracked

Sessions (7d)

43

Cache Hit Rate

69.3%

Avg Duration

17m

Avg Input Tok

1,139,737

Avg Output Tok

15,468

Total Sessions

670

since 2026-02-17

Model Mix (7-day)

opus
100.0%

Session Quality (7-day)

30

Excellent

85+

12

Good

70-84

1

Warning

50-69

0

Poor

<50

Daily Sessions & Cache Hit (14-day)

Weekly Overhead Trend

2026-W12
5,631,235
53s
2026-W13
1,081,343
21s
2026-W14
1,822,687
14s
2026-W15
555,078
11s
2026-W16
1,098,001
9s
2026-W17
1,931,043
24s
2026-W18
601,930
26s

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

Project Breakdown (7-day)

Project Sessions Avg Min Input Tok
rgdm-agency 43 17 49,008,701

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