Vault Synthesizer · concepts written per night · 60 nights
8 consecutive nights
The synthesizer wrote zero concepts for 8 nights running. The vault-synthesizer eval suite caught it on day 9. Recovered 2026-05-11 — the dashboard preserves the incident as proof of operational maturity.
Eval pass · last night
7 / 10
Fleet spend · 30 days
$35.26
92.5% local · 360 runs
Local-only share
92.5%
Most agents · $0/run
Spend governors
$50 / mo
$7 task · $20 daily breaker
vault_indexer
7 runs · 7d
vault_synthesizer
9 runs · 7d
deep_researcher
7 runs · 7d
meta_agent
quiet · 7d
daily_driver
7 runs · 7d
knowledge_lint
1 runs · 7d
flush
46 runs · 7d
job_feed
54 runs · 7d
24-hour activity · 8 agents ·
26 runs
vault_indexer
vault_synthesizer
deep_researcher
meta_agent
daily_driver
knowledge_lint
flush
job_feed
15:41
21:41
03:41
09:41
15:41
30-day cost trend · local vs cloud · stacked
Model mix · last 30 days
- local 92.5% 333 runs
- cloud 7.5% 27 runs
Synthesizer telemetry · concepts written · 60 nights
Eval suite · vault-synthesizer · latest run
- vs-014 output-completeness skipped
- vs-015 output-completeness passed
- vs-016 status-misreport passed
- vs-017 status-misreport passed
- vs-018 schema-integrity passed
- vs-019 config-fail-loud passed
- vs-020 index-integrity passed
- vs-021 downstream-consumer passed
- vs-012 source-attribution skipped
- vs-013 stale-overweighting skipped