A
100
fetch-tweets
7d: 28r · 100% ok30d: 111r · 100% ok● stable
← back to agent dashboard
SKILL HEALTH
Per-skill quality grades computed deterministically from the last 7 and 30 days of pipeline runs. No LLM here — the formula is the interpretation. Reproducible and auditable.
Updated 6/23/2026, 4:43:46 AM
fleet summary
A
5
healthy
B
1
mostly fine
C
0
watch
D
0
action needed
F
0
broken
N/A
0
too few runs
per skill
A
100
narrative-tracker
7d: 28r · 100% ok30d: 111r · 100% ok● stable
A
100
morning-brief
7d: 28r · 100% ok30d: 113r · 100% ok● stable
A
100
digest
7d: 28r · 100% ok30d: 111r · 100% ok● stable
A
100
on-chain-monitor
7d: 28r · 100% ok30d: 111r · 100% ok● stable
B
83
signal-verdict
7d: 8r · 75% ok30d: 29r · 86% ok▼ degrading
success rate dropped